Inspiration as a Psychological Construct

Abstract

Inspiration has received little theoretical or empirical attention within psychology. Inspiration is conceptualized herein as a general construct characterized by evocation, motivation, and transcendence. In Studies 1a and 1b, a trait measure of inspiration was developed and was found to have strong psychometric properties. Studies 2a-2c documented a nomological network consistent with the present conceptualization. Study 3 related inspiration to the holding of U.S. patents. Study 4 linked trait inspiration to daily experiences of inspiration, extended the nomological network to the state level, documented antecedents and consequences, and established incremental validity. This research provides a foundation for further study of inspiration, both as a general construct and in specific content domains (e.g., religion, creativity, interpersonal relations).

Full text

Inspiration as a Psychological Construct
Todd M. Thrash and Andrew J. Elliot
University of Rochester
Inspiration has received little theoretical or empirical attention within psychology. Inspiration is con-
ceptualized herein as a general construct characterized by evocation, motivation, and transcendence. In
Studies 1a and 1b, a trait measure of inspiration was developed and was found to have strong
psychometric properties. Studies 2a–2c documented a nomological network consistent with the present
conceptualization. Study 3 related inspiration to the holding of U.S. patents. Study 4 linked trait
inspiration to daily experiences of inspiration, extended the nomological network to the state level,
documented antecedents and consequences, and established incremental validity. This research provides
a foundation for further study of inspiration, both as a general construct and in specific content domains
(e.g., religion, creativity, interpersonal relations).
Inspiration is an experience with which we are all familiar. We
are inspired when insights or ideas imbue a task with a sense of
necessity and excitement. We are inspired when a mentor or role
model reveals new possibilities that we would not have recognized
on our own. We are inspired when a sense of beauty, truth, or the
divine moves us to pursue a goal more important than the mundane
concerns that often occupy our minds.
Although such experiences of inspiration are no doubt familiar
to the psychologist, the topic has received little sustained attention
within psychology and has been virtually ignored within person-
ality and motivational psychology. Furthermore, inspiration has
typically been conceptualized narrowly within particular content
domains (e.g., religious, creative, interpersonal) or theoretical
frameworks. We advocate a phenomenon-based approach (Stern-
berg & Grigorenko, 2001) to inspiration that both embraces the
breadth of the inspiration concept understood by the layperson and
is informed by diverse theoretical perspectives. Our aims in the
present research are to offer a conceptualization of inspiration, to
validate the inspiration construct, and to establish its importance in
mainstream empirical psychology.
INSPIRATION CONCEPTUALIZED
In its literal sense, inspiration refers to the process of breathing
in or inhaling, but it is the figurative sense that is relevant to
psychology. The first figurative, general definition listed in the
Oxford English Dictionary (OED; Simpson & Weiner, 1989) is the
following: “A breathing in or infusion of some idea, purpose, etc.
into the mind; the suggestion, awakening, or creation of some
feeling or impulse, especially of an exalted kind” (p. 1036). The
concept of inspiration is used in many disciplines, including psy-
chology (Hart, 1998; Kris, 1952; Lockwood & Kunda, 1997;
Taylor & Lobel, 1989), anthropology (Leavitt, 1997), theology
(Canale, 1994a, 1994b), education (Tjas, Nelsen, & Taylor, 1997),
art and literature (Bowra, 1955; Harvey, 1999), management (Bass
& Avolio, 1994; Dess & Picken, 2000), and engineering (Beer,
Quinn, Chiel, & Ritzmann, 1997).
Our reading of the various literatures on inspiration points to
several striking commonalities, thus permitting a general concep-
tualization of the construct: Inspiration implies motivation, which
is to say that it involves the energization and direction of behavior
(Elliot, 1997); inspiration is evoked rather than initiated directly
through an act of will or arising without apparent cause; and
inspiration involves transcendence of the ordinary preoccupations
or limitations of human agency. These characteristics are sug-
gested in the OED definition and have been identified explicitly or
implicitly by theorists across disciplines (Bowra, 1955; Bradley,
1929; Carpenter, 1987; Hart, 1998). We use the term trigger to
refer to the stimulus object that evokes inspiration (e.g., a person
or idea) and target to refer to the object toward which the resulting
motivation is directed (e.g., a possible self, personal goal, or
creative product). In the following, we overview the major ap-
proaches to inspiration and argue that they are compatible with our
conceptualization.
INSPIRATION FROM ABOVE:
SUPERNATURAL SOURCES
In its original usage, inspiration referred to an influence by a
supernatural being in which the individual is used as an instrument
for the delivery of divine truths. Although most modern psychol-
ogists reject the notion of supernatural influence, this account
should be appreciated for what it reveals about inspiration. A
well-known example from ancient Greece is the ascription of
artistic gifts to the influence of a Muse. By whispering into the
poet’s ear, the Muse delivered both divine knowledge and the
ability to communicate it (Leavitt, 1997). It remains common in
modern times for artists, scientists, and other creative individuals
to attribute their best ideas and creative impulses to supernatural,
transcendent, or unknown forces (Ghiselin, 1952; Harding, 1948).
Todd M. Thrash and Andrew J. Elliot, Department of Clinical and
Social Sciences in Psychology, University of Rochester.
We thank Kirk Brown and Jennifer LaGuardia for collecting the data
used in Study 1a. We also thank Ed Deci, Rich Ryan, and Jon Haidt for
their helpful comments on a draft of this article.
Correspondence concerning this article should be addressed to Todd M.
Thrash, Department of Clinical and Social Sciences in Psychology, Human
Motivation Program, Meliora Hall, University of Rochester, Rochester,
New York 14627. E-mail: thrash@psych.rochester.edu
Journal of Personality and Social Psychology Copyright 2003 by the American Psychological Association, Inc.
2003, Vol. 84, No. 4, 871–889 0022-3514/03/$12.00 DOI: 10.1037/0022-3514.84.4.871
871

A related concept of inspiration is found in Judeo-Christian
accounts of the writing of scripture. From this perspective, inspi-
ration is a God-given capacity to transmit the word of God,
whether directly (Aquinas, 1950) or as mediated by the writer’s
experience or interpretation of a divine encounter (Schleiermacher,
1963). In theology, the term revelation refers to the disclosure of
divine knowledge to the individual, whereas inspiration typically
refers to the transmission or translation of the revelation into
written form. A connection between revelation-like experiences
and inspiration has also been emphasized in the area of transper-
sonal psychology (Hart, 2000).
The account of inspiration as supernatural influence provides a
paradigmatic illustration of the characteristics of motivation, evo-
cation, and transcendence. We may regard inspiration from above
as a motivational state evoked by a revelation (trigger) and di-
rected toward the conversion of transcendent, revealed knowledge
into a work of art, a text, or some other concrete form (target).
INSPIRATION FROM WITHIN:
INTRAPSYCHIC SOURCES
With the emergence of the field of psychology toward the end of
the 19th century, theorists sought a scientific account of inspiration
in terms of intrapsychic processes (Preminger, 1965). Some theo-
rists (e.g., Raymond, 1907) aimed to reconcile theology with the
emerging science, but the emphasis in psychology was on replac-
ing supernatural with deterministic explanation (e.g., Kris, 1952;
Ribot, 1906). The majority of work on intrapsychic sources of
inspiration has focused on the creative rather than the religious
domain (for exceptions, see Batson, Schoenrade, & Ventis, 1993;
Fauteux, 1994), and divine revelation has been replaced by illu-
mination (i.e., the insight or eureka experience) from within the
human psyche.
Early approaches in psychology attributed inspiration to the
unconscious (Ribot, 1906; von Hartmann, 1884; Wallas, 1926).
According to von Hartmann (1884), the unconscious produces
ideas that are more organic and elegant than those manufactured
by the will. Wallas (1926) suggested that the creative process
consists of four stages: preparation, incubation, illumination, and
verification. The conscious will dominates during preparation and
verification, whereas unconscious, unwilled processes are respon-
sible for incubation and for producing illumination.
A second source posited to produce inspiration is the precon-
scious (Kris, 1952; Martindale, 1981; Rothenberg, 1990), from
which ideas are more readily accessible. Building on Freudian
concepts, Kris (1952) explained creative inspiration in terms of a
regression in the service of the ego. In the inspiration stage, the
individual permits himself or herself to regress to primary pro-
cesses that involve flexible thinking, thus making novel associa-
tions more likely. Once a promising idea emerges, the elaboration
phase begins—the individual reinstates ego boundaries and uses
secondary processes for reality testing and revision. Empirical
research has confirmed a link between controlled regression and
creativity (for a review, see Suler, 1980).
Finally, some theorists have pointed to the willful, though
typically indirect, generation of novel ideas (Osborn, 1957; Ward,
Finke, & Smith, 1995). Creative cognition researchers have found
that intentionally merging unrelated images may yield preinven-
tive forms that have appealing emergent properties that may be
extended to a new invention (Finke, 1995). Although ideas may be
generated willfully, inspiration tends to be associated less with the
generation of ideas than with their perception post hoc.
These accounts illustrate the characteristics of inspiration that
we posited above. Although the sources of inspiration are intra-
psychic, inspiration is nevertheless evoked in that ideas impinge on
consciousness from the unconscious, the preconscious, or the
perceptual field. Transcendence is illustrated by the fact that the
individual gains access to and uses ideas that are felt to be more
elegant or novel than those generated willfully. These approaches
imply the importance of motivation, focusing on the direction of
motivation by pointing to the ideas or illuminations that guide
creative behavior. Energization has been neglected, perhaps be-
cause of the lasting influence of stage theories (Kris, 1952; Wallas,
1926), which treat illumination as the culmination of unwilled
processes, followed by the labor of verification. Autobiographical
accounts of creators, however, suggest that good ideas may
strongly energize subsequent work (Gnezda-Smith, 1994; Harding,
1948). To illustrate, Donaldson (quoted in Ward, 2001) stated, “As
soon as those two ideas came together, my brain took fire...I
spent the next three months feverishly taking notes, drawing maps,
envisioning characters, studying the implications” (pp. 350, 353).
Inspiration from within may thus be conceptualized as a motiva-
tional state that is triggered by a compelling idea or illumination
(including nonverbal ideas; Sadoski & Paivio, 2001) and that is
targeted toward the actualization or realization of the idea.
INSPIRATION FROM WITHOUT:
ENVIRONMENTAL SOURCES
Recent writings have emphasized sources of inspiration in the
external environment (e.g., people, nature), although external
sources were recognized as early as in ancient Greece. Plato
suggested that poets may inspire their readers, thus transmitting the
Muse’s influence (Rothenberg & Hausman, 1976). Christians see
Christ as a human source of divine inspiration (Schwo¨bel, 1987).
Gilson (1953) pointed to the Muse-like women who inspired
various writers. Composers interviewed by McCutchan (1999)
found inspiration in music, nature, poetry, and other sources.
Others have pointed to the capacity for inspiration from managers,
mentors, role models, and heroes (Dess & Picken, 2000; Jung,
1986; Pleiss & Feldhusen, 1995; Tjas et al., 1997). Animals and
even insects have inspired recent work in robotics (Beer et al.,
1997).
Existential and humanistic approaches to inspiration and related
phenomena have emphasized the importance of openness, vari-
ously conceptualized as allocentric perception (Schachtel, 1959),
B-cognition (Maslow, 1968), encounter (May, 1975), I–thou rela-
tion (Buber, 1996), and openness to experience (Rogers, 1961; see
also McCrae & Costa, 1997). These approaches suggest that ob-
jects in the world are capable of evoking inspiration, if only one
may put aside one’s tendency to perceive them narrowly as objects
of desire, habit, or comparison.
Emotion researchers have begun to examine self-transcendent
emotions (Haidt, 2000, 2003; Haidt & Keltner, in press; Keltner &
Haidt, in press). Haidt (2000, 2003) has provided evidence for the
experience of elevation, a positive emotion that is elicited when
one witnesses virtue and that produces a desire to be virtuous in
turn. Whereas elevation is elicited by virtue, awe is elicited by
872
THRASH AND ELLIOT

vastness, and admiration is elicited by skill (Keltner & Haidt, in
press). These eliciting qualities are reminiscent of the Greek tran-
scendentals of truth, beauty, and goodness, all of which might be
regarded as qualities that inspire.
Several theorists have implicated implicit motives in inspiration
(McAdams, 1982; McClelland & Kirshnit, 1988; Steele, 1977).
McClelland and Kirshnit (1988) found that viewing a film about
Mother Teresa produced a positive immunological response. Indi-
viduals who were dispositionally high in the relaxed affiliative
syndrome (i.e., higher in need for affiliation than in need for power
and low in activity inhibition) were most responsive. Steele (1977)
found that viewing inspirational speeches by Winston Churchill
and others led to increased power motivation and general activa-
tion. Individuals who were dispositionally high in need for power
were most responsive. Thus, different motives may play a role in
different types of inspiration, just as emotion researchers have
identified several inspiration-relevant emotions.
Inspiration has also been investigated by social comparison
theorists (Lockwood & Kunda, 1997, 1999; Taylor & Lobel, 1989;
Wood, 1989). Taylor and Lobel (1989) reviewed evidence that
cancer patients prefer contact with other patients who are doing
better than they are, so that they can gain inspiration and coping
information (see also Helgeson & Taylor, 1993). Lockwood and
Kunda (1997, 1999) found that exposure to high-achieving role
models led participants to adopt more positive self-conceptions
and inspired them to set higher aspirations but that such effects
occur only when the role model’s successes are personally relevant
and attainable.
In these approaches, inspiration is evoked by a person or object
in the external environment. The individual is moved by what is
good or beautiful (Haidt & Keltner, in press; May, 1975; McClel-
land & Kirshnit, 1988) or superior to the self in some way
(Lockwood & Kunda, 1997), which illustrates transcendence.
These approaches recognize the motivational nature of inspiration,
although most have emphasized the energization component by
focusing on responsiveness to inspiring trigger objects. Less at-
tention has been given to target objects and, thus, to whether, and
toward what, the motivation is directed. One exception is social
comparison theory, which has specified not only a trigger object (a
superior individual) but also a target object (a future self). This
precision, however, limits the scope of social comparison theory.
Exposed to a superior other, the aspiring achiever might envision
a better future self (as posited by social comparison theory), but the
writer may envision a protagonist, the artist may envision a new
sculpture, and so on.
INSPIRATION AS A STATE AND TRAIT
In their classic compilation of 17,953 terms descriptive of per-
sonality, Allport and Odbert (1936) classified inspirable as a trait
descriptor, inspired as a temporary state, and inspirational as a
social evaluation. Several studies have examined the inspired state
as it is understood by lay research participants. Davitz (1969)
found that participants characterized inspiration as involving en-
hancement, activation, hyperactivation, comfort, and a feeling of
being overtaken. In his semantic atlas of words connoting emo-
tions, Averill (1975) reported that participants rated inspired above
average in activation, positive evaluation, depth, and uncontrol.
Dahl and Stengel (1978) found the inspired state to be character-
ized by positivity, activation, and attraction emanating from (rather
than to) the object. On the basis of factor analysis, Watson, Clark,
and Tellegen (1988) included “inspired” as an item on their mea-
sure of positive affect. Hart (1998) asked participants to recount
experiences of inspiration. The inspired state was found to be
characterized by feelings of connection, openness, clarity, and
energy.
These findings regarding lay conceptions of the inspired state
are consistent with our portrait of inspiration as involving moti-
vation (e.g., activation, energy), evocation (e.g., feeling overtaken,
uncontrol, attraction from the object, openness), and transcendence
(e.g., positivity, enhancement, clarity). Although Allport and Od-
bert’s (1936) analysis suggests that inspiration may be viewed as
a trait as well as a state, no research has directly examined
inspiration as a trait.
INTEGRATION AND PRESENT RESEARCH
As the above review illustrates, the literature on inspiration has
been fragmented in several respects. First, inspiration has been
studied most often in its specific manifestations—for instance, as
a religious, creative, or interpersonal phenomenon. Such concep-
tual fragmentation is at odds with the folk concept of inspiration
(and the historical tradition from which it emerged), which is
readily applied to phenomena across domains. Second, the various
manifestations of inspiration have been closely associated with
particular theoretical frameworks or disciplines. To understand the
inspiring influence of a person, one might consult social compar-
ison theory, but to understand the inspiring influence of an insight,
one might consult a psychodynamic approach. Such theoretical
fragmentation is at odds with the aim of parsimony.
In an effort to integrate the literature, we have shown that
several themes pervade the various conceptualizations of inspira-
tion. Inspiration is evoked by some trigger stimulus, be it a divine
revelation, a creative illumination, or a person or object in the
external environment. The inspired individual is moved by the
truth, ingenuity, goodness, beauty, or superiority of the trigger
object and is motivated to transmit, actualize, or emulate those
transcendent qualities. Inspiration is thus conceptualized as a broad
construct that spans multiple content domains (e.g., religious,
creative), sources (e.g., intrapsychic, external), triggers (e.g., illu-
mination, nature), transcendent qualities (e.g., beauty, goodness),
and targets (e.g., products, possible selves).
In the present research, inspiration is conceptualized as both a
trait and a state, because it is presumed to vary both between and
within individuals. In Studies 1a and 1b, we develop a trait
measure of inspiration and evaluate its psychometric properties.
We then assess the measure’s construct validity by examining its
nomological network (Studies 2a, 2b, and 2c) and by conducting a
known-groups analysis (Study 3). In Study 4, we use an
experience-sampling methodology to examine predictive validity,
within-person correlates, antecedents and consequences, and in-
cremental validity.
STUDY 1: DEVELOPMENT OF THE INSPIRATION
SCALE AND PSYCHOMETRIC PROPERTIES
The purpose of Study 1 is to develop a trait measure of inspi-
ration and to evaluate its psychometric properties in two samples
873
INSPIRATION

(Studies 1a and 1b). Because we seek to understand a concept that
is meaningful in individuals’ lives (Hart, 1998) but underappreci-
ated by psychologists, we aim to develop a measure that is straight-
forward and face valid. Research attests to the utility of such
measures in the study of important psychological constructs, such
as life satisfaction (Andrews & Withey, 1976), self-esteem (Rob-
ins, Hendin, & Trzesniewski, 2001), and gratitude (McCullough,
Emmons, & Tsang, 2002).
Study 1a
Method
Preliminary Instrument Development
In pilot research, we drafted 19 inspiration items, with response options
ranging from 1 (strongly disagree)to7(strongly agree), and we admin-
istered the items to a sample of 181 undergraduates. A promising set of 4
items was selected on the basis of item content (e.g., balance regarding
references to triggers and targets, and inclusion of only domain-general
items) and on the basis of psychometric criteria (internal consistency and
unifactorial structure in an exploratory factor analysis [EFA]). To better
capture the dynamic nature of inspiration, we adapted these items to consist
of four statements, each of which is rated in terms of both frequency and
intensity (see the Appendix). Frequency items are rated on a scale from 1
(never)to7(very often), and intensity items are rated from 1 (not at all)
to7(very deeply or strongly).
1
Participants and Procedure
A total of 333 (119 male and 214 female) undergraduates completed the
inspiration measure in small-group sessions in return for extra credit in a
psychology course.
Results and Discussion
Factor Structure
We examined the factor structure of the measure using confir-
matory factor analysis (CFA).
2
Analysis of the covariance matrix
was conducted using Amos 4.01 (Arbuckle, 1999) with maximum
likelihood estimation. The four Frequency and four Intensity items
were specified to be indicators of frequency and intensity latent
variables, respectively. The uniquenesses of paired items (e.g.,
Items 1f and 1i in the Appendix) were permitted to covary. We
identified the model by constraining latent variable variances to 1.
This model was found to have an acceptable fit (see Model 1 in
Table 1). All parameters were significant, and all standardized
loadings exceeded .80. The two latent variables were strongly
correlated (r ⫽ .81, p ⬍ .0001); nevertheless, the two-factor model
fit better than a one-factor model, ⌬
␹
2
(1, N ⫽ 333) ⫽ 355.93, p ⬍
.001.
Internal Consistency
We formed inspiration frequency and intensity indices by sum-
ming the respective sets of four items, and we formed an overall
index by summing all eight items. All three indices were found to
be internally consistent (see Table 2 for reliabilities and descriptive
statistics).
These results indicate that the new measure, hereafter called the
Inspiration Scale (IS), is internally consistent and consists of
distinguishable and internally consistent Frequency and Intensity
subscales. In Study 1b, we aim to replicate these results and
demonstrate temporal stability.
Study 1b
Method
A total of 220 (86 male and 134 female) undergraduates participated in
the study in return for extra credit in a psychology course. All participants
completed the IS at Time 1, and 212 (81 male and 131 female) participants
completed the measure again 7.5 weeks later. Participants received ques-
tionnaires in class and completed them outside of class. An overall inspi-
ration index and frequency and intensity indices were formed as in
Study 1a.
Results and Discussion
Factor Structure
We tested a CFA model identical to that examined in Study 1a
using the Time 1 data. The model was found to have an acceptable
fit (see Model 2 in Table 1). All parameters were significant, and
all loadings exceeded .70. The frequency and intensity latent
variables were highly correlated (r ⫽ .60, p ⬍ .001). These results
replicate the structure documented in Study 1a.
Internal Consistency
As in Study 1a, the IS and its Frequency and Intensity subscales
were each found to be internally consistent (see Table 2).
1
The four statements that we selected are sufficiently general to span the
various manifestations of inspiration (i.e., all combinations of domains,
sources, triggers, targets, transcendent qualities, etc.). Alternatively, we
could have attempted to include separate, specific items for all possible
manifestations of the construct. We opted against this latter approach for
two reasons. First, as was apparent during our initial generation of items,
the nascent state of the literature precluded the possibility of evenly and
nonarbitrarily sampling items across the content universe of the construct.
Second, we presumed that what is most important is the extent to which
individuals experience inspiration, not their mean level of inspiration
across all of its conceivable phenotypic manifestations. We would not
suggest, for instance, that a religious saint is low in inspiration because he
or she has never been inspired to paint or has few role models. Our
approach is akin to the assessment of motives, genotypic constructs that are
manifest in diverse ways in the lives of individuals (Thrash & Elliot, 2001).
Our approach also resembles the assessment of global self-esteem (Robins
et al., 2001; Rosenberg, 1965), an important construct that is not equivalent
to mean esteem across specific content domains (Harter, 1993; Marsh,
1986).
2
In establishing the factor structure of a scale, researchers often conduct
an EFA in an initial sample, followed by a CFA in an independent sample.
We did not use EFA to examine the new eight-item measure, because EFA
assumes that uniquenesses are uncorrelated, an assumption that is not
tenable in the present context (see the model specification). Hoyle (1991)
discussed the advantages of CFA over EFA for a model identical in
structure to ours. For the sake of demonstrating replicability, a second CFA
with an independent sample is conducted in Study 1b.
874
THRASH AND ELLIOT

Stability Across Time
The concept of stability may be applied to (a) measurement
properties of an instrument (e.g., model structure, loadings), (b)
construct parameters (e.g., means, variances), and (c) the relative
ordering of individuals (test–retest correlations). These aspects of
stability were examined using means-and-covariance-structure
(MACS) analysis (Vandenberg & Lance, 2000).
Measurement invariance. We examined stability of the model
structure by specifying a four-factor model that incorporated
Time 1 and Time 2 models simultaneously. The Time 1 and
Time 2 portions of the model were both identical to the two-factor
model examined above. Covariances were specified among all four
latent variables and between uniquenesses of identical items at
Times 1 and 2. We identified the model by constraining one
loading on each latent variable to 1 and the corresponding inter-
cepts to 0. This model was found to have an acceptable fit (see
Model 3 in Table 1). All latent variable variances/covariances and
factor loadings were significant, and all loadings exceeded .70 (see
Figure 1). Having established stability of the model structure, we
cumulatively imposed equality constraints that forced correspond-
ing parameters at Times 1 and 2 to be equal and assessed whether
the constraints produced decrements in fit. As shown in Table 1,
constraining factor loadings (Model 4) and intercepts (Model 5) to
be invariant produced no decrement in fit. The IS thus demon-
strated stable measurement properties across time.
Stability of construct parameters. Further constraining the la-
tent frequency and intensity variances (Model 6), covariance
(Model 7), and means (Model 8) to be invariant across time
produced no decrement in model fit. These results establish sta-
bility of the construct-level parameters of the frequency–intensity
model.
Stability of individual differences. Test–retest correlations for
latent frequency and intensity in Model 8 were both .77, indicating
that individual differences were quite stable.
In sum, Study 1 finds, first, that the IS consists of two internally
consistent factors, justifying the use of separate four-item fre-
quency and intensity indices. Second, the factors were strongly
correlated, and the overall scale was internally consistent, thus also
justifying use of the overall eight-item index. Third, the measure-
ment properties of the IS were invariant across time, the construct
parameters were stable, and individual differences were also sta-
ble, indicating that the IS is suitable as a trait measure. The strong
psychometric properties of the IS permit a shift in Studies 2 and 3
to the substantive issue of construct validity. In Study 2, we
examine the nomological network of inspiration in three samples
(Studies 2a, 2b, and 2c).
Table 1
Study 1: Fit Indices for CFAs and Tests of Stability Across Time
Model
␹
2
df CFI TLI RMSEA ⌬
␹
2
⌬df ⌬CFI
Two-factor CFAs
1. Study 1a 38.12** 15 .99 .98 .07
2. Study 1b, Time 1 33.91** 15 .99 .97 .08
Tests of measurement invariance
(Study 1b)
3. Baseline four-factor model 141.64*** 82 .98 .97 .06
4. Invariant loadings 145.58*** 88 .98 .97 .06
Model 4 vs. 3 3.94 6 .00
5. Invariant intercepts 148.48*** 94 .98 .98 .05
Model 5 vs. 4 2.90 6 .00
Tests of stability of construct
parameters (Study 1b)
6. Invariant latent variances 153.27*** 96 .98 .98 .05
Model 6 vs. 5 4.79 2 .00
7. Invariant latent covariance 153.56*** 97 .98 .98 .05
Model 7 vs. 6 .29 1 .00
8. Invariant latent means 158.23*** 99 .98 .98 .05
Model 8 vs. 7 4.67 2 .00
Note. Constraints were added cumulatively to previously accepted models. The accepted model in each model
comparison is indicated with italics. For the chi-square in Model 1, N ⫽ 333; for Models 2–8, N ⫽ 220. CFA ⫽
confirmatory factor analysis; CFI ⫽ comparative fit index; TLI ⫽ Tucker–Lewis index; RMSEA ⫽ root-mean-
square error of approximation.
** p ⬍ .01. *** p ⬍ .001.
Table 2
Study 1: Descriptive Statistics and Reliabilities of the
Inspiration Scale
Variable MSDRange Cronbach’s
␣
Study 1a
Overall index
a
35.29 10.05 8–56 .95
Frequency
a
17.39 5.46 4–28 .93
Intensity
a
17.90 5.21 4–28 .92
Study 1b, Time 1
Overall index
b
38.38 8.14 8–56 .91
Frequency
c
18.61 4.50 4–28 .90
Intensity
b
19.75 4.69 4–28 .92
Study 1b, Time 2
Overall index
d
38.94 7.55 10–56 .90
Frequency
d
19.08 4.33 5–28 .90
Intensity
d
19.86 4.31 5–28 .91
a
n ⫽ 333.
b
n ⫽ 219.
c
n ⫽ 220.
d
n ⫽ 212.
875
INSPIRATION

STUDY 2: NOMOLOGICAL NETWORK
OF INSPIRATION
Study 2a
In Study 2a, we examine the relationship between inspiration
and variables representing three major theoretical frameworks in
personality and motivational psychology: approach–avoidance
motivation (e.g., Elliot & Thrash, 2002), intrinsic–extrinsic moti-
vation (e.g., Deci & Ryan, 1985), and the Big Five traits (e.g.,
McCrae & Costa, 1987). Given that inspiration has been concep-
tualized in terms of positive qualities to be approached, it was
hypothesized to be positively related to the behavioral activation
system (BAS), an approach system, and unrelated to the behavioral
inhibition system (BIS), an avoidance system (Gray, 1987). Inspi-
ration was expected to be positively related to intrinsic motivation,
because both are focused on incentives inherent to a task, and
unrelated or negatively related to extrinsic motivation. Regarding
the Big Five, inspiration was hypothesized to be positively related
to Openness to Experience, which may facilitate receptiveness to
evocative stimuli. Inspiration was also expected to be positively
related to Extraversion but unrelated to Neuroticism, given evi-
dence that the core of these traits represent approach and avoid-
ance dispositions, respectively (Elliot & Thrash, 2002; Lucas,
Diener, Grob, Suh, & Shao, 2000). Inspiration was not expected to
relate to Agreeableness or Conscientiousness. We additionally
examined relationships with sex, grade point average (GPA), and
students’ university majors. All hypotheses were undifferentiated
with respect to the frequency and intensity dimensions of
inspiration.
Method
Participants and Procedure
A total of 152 (51 male and 101 female) undergraduates completed
questionnaire packets in small-group sessions in return for extra credit in a
psychology course.
Measures
Inspiration. The IS was used to assess inspiration frequency and in-
tensity and overall inspiration.
BAS and BIS. Carver and White’s (1994) measure was used to assess
BAS (13 items) and BIS (7 items). Items were rated on a scale from 1
(strongly disagree)to4(strongly agree).
Intrinsic and extrinsic motivation. Amabile, Hill, Hennessey, and
Tighe’s (1994) Work Preference Inventory was used to assess individual
differences in intrinsic motivation (15 items) and extrinsic motivation (15
items). Items were rated on a scale from 1 (never or almost never true of
me)to4(always or almost always true of me).
Big Five traits. Costa and McCrae’s (1992) NEO Five Factor Inven-
tory was used to assess Openness to Experience, Extraversion, Neuroti-
cism, Agreeableness, and Conscientiousness (12 items per trait). Items
were rated on a scale from 1 (strongly disagree)to5(strongly agree).
Sex, GPA, and majors. Participants recorded their sex (coded
male ⫽ 1, female ⫽ 2), GPA, and majors. The number of majors listed
formed a number-of-majors variable (undecided ⫽ 0, single major ⫽ 1,
double major ⫽ 2). For individuals with at least one major, three dummy-
coded variables were used to represent whether or not (1 or 0) the
individual was pursuing a major within the humanities, social sciences, or
natural sciences.
Results and Discussion
Descriptive statistics for each variable may be found in Table 3,
and correlations between inspiration and other variables may be
found in Table 4. As expected, the overall inspiration index cor-
related positively with BAS (but not BIS), intrinsic motivation (but
not extrinsic motivation), and Openness to Experience and Extra-
version (but not Neuroticism, Agreeableness, or Conscientious-
ness). Inspiration thus converged with several measures of moti-
vation, but only with those that are appetitive or task focused. The
fact that inspiration correlated with Openness to Experience but
not Conscientiousness is consistent with our conceptualization of
inspiration as evoked and unwilled. The overall inspiration index
was positively related to the number of majors pursued. This
relationship reflects both low inspiration among individuals who
were undecided about a major and high inspiration among double
majors (undecided, M ⫽ 32.9; single major, M ⫽ 36.9; double
major, M ⫽ 39.4) and suggests that inspired individuals are more
engaged with their environment. Individuals high in inspiration
were most likely to be pursuing majors within the humanities,
which include the disciplines most often associated with inspira-
tion (e.g., art, religion). The overall inspiration index was unrelated
to sex and GPA. Although the frequency and intensity dimensions
demonstrated slightly different patterns of significant correlations,
the differences in magnitude were negligible in most cases.
Figure 1. Study 1b: Longitudinal model illustrating the frequency–
intensity structure of the Inspiration Scale at Times 1 and 2. For clarity of
presentation, uniquenesses and uniqueness covariances are not illustrated.
Coefficients not enclosed in parentheses are from the baseline model
(Model 3), and coefficients in parentheses are from the final model (Model
8). Parameters are presented in standardized form and are highly signifi-
cant. Standardized parameters for Model 8 are not necessarily equal at
Times 1 and 2, because invariance tests are conducted on unstandardized
parameters. Items may be found in the Appendix.
876
THRASH AND ELLIOT

Regressing the overall inspiration index simultaneously on the
Big Five traits revealed that inspiration was uniquely related to
Openness to Experience, F(1, 146) ⫽ 30.64, p ⬍ .001 (
␤
⫽ .42),
and (marginally) to Extraversion, F(1, 146) ⫽ 3.09, p ⬍ .10 (
␤
⫽
.15). The multiple R
2
was .22, indicating that only about one fifth
of the variance of the IS overlaps with the Big Five traits.
Study 2b
In Study 2b, we aim to extend inspiration’s nomological net-
work. Unlike states that lack object directedness or intentionality
(e.g., moods; Frijda, 1994), inspiration has been conceptualized as
focused on transcendent stimulus qualities. Accordingly, we ex-
pected inspiration to be positively related to trait absorption (Tel-
legen, 1981), a readiness to enter psychological states involving
engrossed attention to stimulus qualities such as beauty. Epstein
(1994) has described two modes of cognitive processing: experi-
ential processing, which is intuitive and concrete, and rational
processing, which is rational and abstract. Inspiration was ex-
pected to be positively related to experiential processing, given
that inspiring stimuli tend to reach consciousness through experi-
ential and perceptual processes (e.g., illumination). Inspiration was
also expected to be positively related to rational processing, given
that inspiring influences are not merely appreciated experientially;
they exemplify abstract, transcendent values toward which the
individual becomes oriented. We also expected inspiration to be
positively related to the work-mastery component of need for
Table 3
Study 2: Descriptive Statistics and Reliabilities
Variable MSDRange Cronbach’s
␣
Study 2a
a
Inspiration (overall) 36.89 8.01 8–56 .92
Inspiration frequency 18.32 4.54 4–28 .91
Inspiration intensity 18.57 4.30 4–28 .89
BAS 42.41 4.69 27–51 .79
BIS 21.18 4.02 9–28 .79
Intrinsic motivation 46.14 5.76 22–57 .80
Extrinsic motivation 40.91 6.93 22–55 .82
Openness to Experience 42.61 6.50 23–57 .71
Extraversion 44.10 6.23 29–56 .79
Neuroticism 32.07 7.80 15–55 .83
Agreeableness 46.02 6.09 30–57 .76
Conscientiousness 44.52 7.63 20–58 .83
No. majors 1.07 0.51 0–2
Major in humanities 0.15 0.36 0–1
Major in social sciences 0.61 0.49 0–1
Major in natural sciences 0.35 0.48 0–1
Sex 1.66 0.47 1–2
GPA 3.30 0.39 2.09–4.00
Study 2b
b
Inspiration (overall) 38.35 9.26 8–56 .94
Inspiration frequency 19.04 5.08 4–28 .93
Inspiration intensity 19.31 4.89 4–28 .92
Absorption 55.93 7.06 36–68 .89
Experiential processing 73.18 10.56 47–98 .88
Rational processing 76.00 11.98 25–98 .90
Work mastery 52.24 6.90 30–68 .80
Competitiveness 16.02 4.52 6–25 .82
Fear of failure 22.76 6.75 9–45 .85
Creativity 30.07 5.04 14–40 .82
Study 2c
c
Inspiration (overall) 39.26 8.52 17–56 .93
Inspiration frequency 19.26 4.54 8–28 .92
Inspiration intensity 20.00 4.68 8–28 .92
Positive affect 36.30 5.72 24–49 .81
Positive affect without
“inspired” item
32.93 5.21 20–44 .81
Negative affect 22.51 6.87 10–41 .89
Positive emotionality 20.83 3.52 11–25 .88
Negative emotionality 13.23 4.52 5–24 .85
Perceived competence 34.03 6.22 18–45 .89
Self-esteem 32.49 5.58 18–40 .87
Optimism 21.34 5.37 8–30 .89
Self-determination 64.04 13.61 32–90 .88
OED inspiration frequency 4.33 1.48 1–7
OED inspiration intensity 4.47 1.55 1–7
Social desirability 47.84 5.24 34–61 .77
Note. BAS ⫽ behavioral activation system; BIS ⫽ behavioral inhibition
system; GPA ⫽ grade point average; OED ⫽ Oxford English Dictionary.
a
n ⫽ 152 except for the three major content area variables (n ⫽
136).
b
n ⫽ 161.
c
n ⫽ 99.
Table 4
Study 2: Correlates of the Inspiration Scale and Subscales
Individual-differences variable
Inspiration
Overall Frequency Intensity
Study 2a
a
BAS .18* .16* .16
BIS ⫺.07 ⫺.01 ⫺.12
Intrinsic motivation .43*** .43*** .35***
Extrinsic motivation ⫺.17* ⫺.15 ⫺.16*
Openness to Experience .43*** .36*** .42***
Extraversion .20* .24** .12
Neuroticism ⫺.11 ⫺.14 ⫺.06
Agreeableness .14 .15 .10
Conscientiousness .02 .11 ⫺.08
No. majors .20* .21** .14
Major in humanities .21* .23** .15
Major in social sciences ⫺.01 .04 ⫺.05
Major in natural sciences ⫺.12 ⫺.15 ⫺.06
Sex ⫺.06 .03 ⫺.13
GPA .00 .00 .00
Study 2b
b
Absorption .48*** .43*** .47***
Experiential processing .24** .18* .27**
Rational processing .22** .22** .18*
Work mastery .33*** .33*** .28***
Competitiveness ⫺.17* ⫺.15 ⫺.17*
Fear of failure ⫺.16* ⫺.20* ⫺.09
Creativity .37*** .39*** .29***
Study 2c
c
Positive affect .58*** .56*** .52***
Positive affect, without “inspired”
item
.49*** .47*** .43***
Negative affect .02 .02 .01
Positive emotionality .26* .30** .18
Negative emotionality ⫺.03 ⫺.01 ⫺.03
Perceived competence .38*** .34** .36***
Self-esteem .33** .34** .28**
Optimism .35*** .36*** .28**
Self-determination .34** .30** .33**
OED inspiration frequency .59*** .57*** .53***
OED inspiration intensity .50*** .33** .60***
Social desirability .16 .18 .12
Note. BAS ⫽ behavioral activation system; BIS ⫽ behavioral inhibition
system; GPA ⫽ grade point average; OED ⫽ Oxford English Dictionary.
a
n ⫽ 152.
b
n ⫽ 161.
c
n ⫽ 99.
* p ⬍ .05. ** p ⬍ .01. *** p ⬍ .001.
877
INSPIRATION

achievement, with which it shares a task focus, but unrelated or
negatively related to the competitiveness component and to the
avoidance-based fear of failure. Finally, we expected that inspira-
tion would relate to creativity, as has been posited since the time
of Plato but has never been tested empirically.
Method
Participants and Procedure
A total of 161 (54 male and 107 female) undergraduates completed
questionnaire packets in small-group sessions in return for extra credit in a
psychology course.
Measures
Inspiration. The IS was used to assess inspiration.
Absorption. Absorption was assessed by the Absorption scale from
Tellegen’s (1982) Multidimensional Personality Questionnaire. This scale
consists of 34 true–false statements.
Experiential and rational processing. Pacini and Epstein’s (1999)
Rational–Experiential Inventory was used to assess experiential (20 items)
and rational (20 items) information processing styles. Items were rated
from 1 (definitely not true of myself)to5(definitely true of myself).
Need for achievement. Spence and Helmreich’s (1983) Work and
Family Orientation Scale was used to assess the work-mastery (14 items)
and competitiveness (5 items) components of the need for achievement.
Items were rated from 1 (strongly disagree)to5(strongly agree).
Fear of failure. Fear of failure was assessed with the 9-item version of
Herman’s (1990) 27-item measure developed by Elliot and Church (2001).
The short form correlates strongly with, covers the same content universe
as, and displays internal consistency and predictive validity comparable to
the full measure. Items were rated from 1 (strongly disagree)to5(strongly
agree).
Creativity. Creativity was assessed with the Problem Solving/Creativ-
ity scale of Marsh and O’Neill’s (1984) Self Description Questionnaire III.
This scale consists of 10 items that are rated from 1 (strongly disagree)to4
(strongly agree).
Results and Discussion
Descriptive statistics and correlates may be found in Tables 3
and 4, respectively. All hypotheses were supported. The overall
inspiration index correlated positively with absorption, consistent
with our conceptualization of inspiration as focused on transcen-
dent qualities of objects. Inspiration correlated positively with both
rational and experiential processing, suggesting that inspiration
engages the head as well as the heart (Epstein, 1994). Inspiration
correlated positively with work mastery but negatively (and more
modestly) with competitiveness and fear of failure. Thus, within
the achievement domain, inspiration converged only with the
motive construct representing appetitive, task-focused engage-
ment, consistent with the results of Study 2a. Inspiration also
correlated positively with creativity, a capacity that has long been
ascribed to inspiration rather than (or in conjunction with) personal
initiative. As in Study 2a, the frequency and intensity indices
related to the other variables in a similar manner.
Study 2c
Study 2c further examines the nomological network of inspira-
tion. Given that studies have found inspiration to be characterized
in part by positivity and arousal (see the introduction), we expected
inspiration to be related to trait positive affect (of which positivity
and arousal are core characteristics; Watson, Wiese, Vaidya, &
Tellegen, 1999) but not trait negative affect. Regarding broader
temperamental styles, inspiration was expected to be related to
positive emotionality but not negative emotionality. Our concep-
tualization of inspiration as evoked, motivational, and transcendent
suggests an influx or bolstering of psychological resources. We
therefore expected inspiration to be positively related to perceived
competence, self-esteem, and optimism. We also expected inspi-
ration to be positively related to self-determination, because in-
spiring influences are likely to be regarded as self-congruent. In
the absence of previous inspiration measures against which to
validate the IS, we also aimed to demonstrate convergence with a
measure based on the OED definition of inspiration. Finally, we
examined whether the IS is related to social desirability.
Method
Participants and Procedure
A total of 99 (58 male and 41 female) undergraduates from various
majors completed questionnaire packets at home in return for $2.
Measures
Inspiration. Inspiration was assessed with the IS and the following
unlabeled definition of inspiration: “A breathing in or infusion of some
idea, purpose, etc. into the mind; the suggestion, awakening, or creation of
some feeling or impulse, especially of an exalted kind” (the first figurative,
general definition of inspiration from the OED; Simpson & Weiner, 1989,
p. 1036). This statement was followed by a frequency item, “How often do
you experience this?” that was rated on a scale from 1 (never)to7(very
often), and an intensity item, “How deeply or strongly do you experience
this (in general)?” that was rated on a scale from 1 (not at all)to7(very
deeply or strongly).
Positive and negative affect. The Positive and Negative Affect Sched-
ule (PANAS; Watson et al., 1988) was used to assess positive affect (10
items) and negative affect (10 items). Participants rated each item on a
scale from 1 (very slightly or not at all)to5(extremely) regarding how they
generally feel.
Positive and negative emotionality. Wills, Windle, and Cleary’s
(1998) measures were used to assess positive emotionality (five items) and
negative emotionality (five items). Items were rated on a scale from 1 (not
at all true)to5(very true).
Perceived competence. Perceived competence was assessed with
O’Brien and Epstein’s (1988) Multidimensional Self-Esteem Inventory.
This measure consists of three statements that are rated on a scale from 1
(strongly disagree)to5(strongly agree) and six questions that are rated on
a scale from 1 (very seldom)to5(very often).
Self-esteem. Self-esteem was assessed using Rosenberg’s (1965) 10-
item Self-Esteem Scale. Items were rated on a scale from 1 (strongly
disagree)to4(strongly agree).
Optimism. Optimism was assessed using Scheier, Carver, and Bridg-
es’s (1994) six-item Life Orientation Test—Revised. Items were rated
from 1 (strongly disagree)to5(strongly agree).
Self-determination. Self-determination was assessed by Sheldon and
Deci’s (1996) 10-item Self-Determination Scale. For each item, partici-
pants rated which of two statements is more true for them. Items were rated
from 1 (only A feels true)to9(only B feels true).
Social desirability. The Marlowe–Crowne Social Desirability Scale
(Crowne & Marlowe, 1964) was used to assess social desirability. This
scale consists of 33 true–false statements.
878
THRASH AND ELLIOT

Results and Discussion
Descriptive statistics and correlates may be found in Tables 3
and 4, respectively. All hypotheses were supported. The overall
inspiration index correlated positively with positive affect and
positive emotionality and was unrelated to negative affect and
negative emotionality, consistent with our conceptualization of
inspiration as appetitive. Inspiration was also positively related to
perceived competence, self-esteem, and optimism, variables that
represent psychological resources. Inspiration was positively re-
lated to self-determination, which suggests that inspiration is com-
patible with the self, even if it is not directly self-instigated. The IS
correlated strongly with the OED measure, which thus grounds the
IS in a culturally shared conception of inspiration and rules out the
possibility that each respondent interprets the IS items in an
idiosyncratic way. Finally, inspiration was unrelated to social
desirability. As in Studies 2a and 2b, inspiration frequency and
intensity had essentially identical nomological networks.
CFA was used to determine whether the IS and OED items
converge as indicators of inspiration and whether inspiration may
be discriminated from positive affect, the strongest correlate of the
IS aside from the OED. The inspiration portion of the model was
identical to the model in Study 1a, except that the OED frequency
and intensity items were included as indicators of the correspond-
ing latent variables (with correlated uniquenesses). The PANAS
Positive Affect items were specified to load on a positive affect
latent variable. The PANAS item “inspired” was also permitted to
load on the inspiration factors. This three-factor model was found
to have acceptable fit,
␹
2
(160, N ⫽ 99) ⫽ 261.54 (comparative fit
index [CFI] ⫽ .92, Tucker–Lewis index [TLI] ⫽ .90, root-mean-
square error of approximation [RMSEA] ⫽ .08). All IS and OED
items had loadings of at least .60, and all PANAS items except
“inspired” had loadings of at least .40. “Inspired” loaded on both
inspiration frequency (.53) and intensity (.33) but not on positive
affect (.00). We trimmed this item’s loading on positive affect,
which resulted in no loss of fit. (Dropping this item from the
observed positive affect variable resulted in a more moderate
correlation with the IS; see Table 4.) This three-factor model fit
better than the two-factor models that we formed by collapsing
latent positive affect with latent inspiration frequency, ⌬
␹
2
(1, N ⫽
99) ⫽ 98.25, p ⬍ .001, or with latent inspiration intensity, ⌬
␹
2
(1,
N ⫽ 99) ⫽ 108.03, p ⬍ .001. The data thus support the conver-
gence of the IS and OED measure of inspiration as well as the
discrimination of inspiration from positive affect.
In sum, Study 2 demonstrates that the IS relates to other mea-
sures in a way that is meaningful, differentiated, and consistent
with our conceptualization. The Frequency and Intensity subscales
were found to have similar patterns of correlations with other
variables, at least within the domain of personality examined
herein.
STUDY 3: KNOWN-GROUPS ANALYSIS
The purpose of Study 3 is to further establish construct validity
by conducting a known-groups analysis. A group that is likely to
be highly inspired is U.S. patent holders; these individuals are
moved by novel ideas and are motivated to translate their ideas
into reality. We hypothesized that patent holders would be inspired
more frequently and intensely than would a sample of university
alumni, a comparison group that has also demonstrated career-
oriented competence yet that we have no reason to believe is
particularly inspired.
Most U.S. patents fall into one of two categories. A utility patent
may be granted to an individual who “invents or discovers any new
and useful process, machine, article of manufacture, or composi-
tions of matter, or any new useful improvement thereof,” whereas
a design patent may be granted to an individual who “invents a
new, original, and ornamental design for an article of manufac-
ture” (United States Patent and Trademark Office [USPTO], 1999,
p. 2). Although utility and design patentees were not expected to
differ in inspiration, we included separate utility and design sam-
ples for exploratory purposes.
It cannot be taken for granted that a measure assesses compa-
rable constructs across populations (Bollen, 1989; Little, 1997). If,
for instance, the IS items are better indicators of inspiration for
patentees than for alumni, then group differences might reflect
differential validity rather than true mean differences. Accord-
ingly, we examined measurement invariance across groups prior to
examining means.
A secondary purpose of this study is to test the hypothesis that,
among patent holders, the number of patents held relates to the
frequency but not the intensity of inspiration.
Method
Participants and Procedure
The USPTO’s CD-ROM-based patentee-assignee database was used to
generate a list of patent holders who received patents in the fourth quarter
of 1999 (1 year prior to data collection), who resided in the United States,
and who were classified as independent inventors. From this list, 150 utility
patentees and 150 design patentees were selected randomly. The University
of Rochester alumni office provided a list of 150 alumni that was generated
randomly, except that it was matched to the patentee list in gender com-
position and was limited to individuals residing in the United States.
Questionnaires were mailed to all 450 individuals. Three weeks later, a
second mailing was sent to individuals who had not responded. A reminder
postcard was sent 4 weeks after the initial mailing, followed by a final
mailing (by certified mail) after 7 weeks.
Sixty-two of the individuals (61 patentees, 1 alumnus) to whom we
mailed materials were unavailable because their addresses were no longer
valid, and we were informed by family members that 4 others were
deceased or out of the country. Additionally, 19 of the final certified
mailings were returned as unclaimed. The final sample of participants who
returned the questionnaire consisted of 51 (47 male, 4 female) utility
patentees, 55 (42 male, 13 female) design patentees, and 93 (84 male, 9
female) alumni, representing 55% of the available sample.
Measures
Inspiration
Inspiration was assessed by the IS.
Number of Patents
The Web-based USPTO database was used to determine the number of
patents held by each patentee. The number of patents ranged from 1 to 93,
with a mean of 6.14 (SD ⫽ 12.50). The distribution was severely skewed
(z ⫽ 19.24, p ⬍ .0001), such that the minimum value (1 patent) was the
modal value (n ⫽ 47). An inverse transformation was used, as recom-
879
INSPIRATION

mended by Tabachnick and Fidell (2001) for L-shaped distributions. This
transformation eliminated skewness (z ⫽ 0.29, ns).
Demographic Variables
Participants were asked to report their sex (coded male ⫽ 1, female ⫽
2), age, and education level. Education level was assessed in terms of the
highest degree attained, and responses were coded on a scale from 1 (none)
to7(doctorate).
Results and Discussion
Measurement Invariance Across Groups
Table 5 displays group means and descriptive statistics. A
two-factor model identical to that examined in Study 1a was
specified in a three-group MACS analysis. We identified the
model by constraining the first loading on each latent variable to
be equal across groups, constraining the corresponding intercepts
to be equal across groups, and standardizing both latent variables
in the alumni group. As shown in Table 6, this baseline model
(Model 1) was found to have an acceptable fit. All latent variable
variances/covariances and factor loadings were significant, and
all loadings exceeded .70 in each group. Constraining loadings
(Model 2) and intercepts (Model 3) to be invariant across groups
produced no decrement in fit. These results establish measurement
invariance and, thus, construct comparability across groups.
Group Differences in Inspiration
Further constraining the latent frequency and intensity variances
(Model 4) and covariance (Model 5) to be invariant produced no
decrement in fit. Such invariance is not necessary for examining
means but is considered elegant nonetheless (Widaman & Reise,
1997).
Constraining latent means to be equal across the utility patent
and design patent groups (Model 6) produced no decrement in fit.
However, as expected, constraining the alumni means to equal the
patentee means (Model 7) produced a decrement in fit due to
sizable group differences in the expected direction for both latent
inspiration frequency (z ⫽ 4.97, p ⬍ .000001) and intensity
(z ⫽ 4.81, p ⬍ .00001).
Number of Patents Held and Ancillary Analyses
Among patent holders, the number of patents held was related to
inspiration frequency (r ⫽ .21, p ⬍ .05) and unrelated to inspira-
tion intensity (r ⫽ .07, ns). Results were similar when group status
(utility vs. design) was controlled (partial r ⫽ .22, p ⬍ .05, and
partial r ⫽ .08, ns, respectively).
We conducted ancillary regression analyses to determine
whether the findings of this study remained robust when we
controlled for sex, age, and education level. The ancillary analyses
yielded results that were essentially identical to those reported
above.
In sum, patent holders were found to be inspired more fre-
quently and more intensely than was a comparison group, which
supports the construct validity of the IS. Among patent holders, the
number of patents held was related to the frequency but not the
intensity of inspiration, thus providing a meaningful discrimina-
tion of these dimensions.
STUDY 4: INSPIRATION IN DAILY EXPERIENCE
In Studies 1–3, we sought to understand the nature of the
inspiration construct by examining trait inspiration. In Study 4, we
assess inspiration and other constructs both as traits and as daily
states using an experience sampling methodology. Our aims are as
follows: to demonstrate that IS trait reports predict individuals’
daily experiences of inspiration, to replicate central components of
the nomological network and to extend this network from the trait
(between-persons) level to the state (within-person) level, to doc-
ument antecedents and consequences across time, and to examine
whether our results are attributable to the variance that inspiration
shares with positive affect.
Method
Participants and Procedure
A total of 171 (58 male and 113 female) undergraduates in an introduc-
tory psychology class completed trait measures at home for extra credit.
Three weeks later, 156 of the individuals participated in a 2-week daily
diary procedure. Six participants were excluded from the diary analyses—5
who dropped out (each by Day 5), and 1 who failed to follow directions—
which resulted in a sample of 150 (48 male and 102 female) in the diary
portion of the study.
The diary consisted of a set of questions that was E-mailed to partici-
pants each day at 5:00 PM for 14 consecutive days. Participants were asked
to reply each night by 2:00 AM or as soon as possible thereafter. Timely
responses earned entries into a lottery for prizes. Items concerned partic-
ipants’ experiences in the roughly 24-hr period since their previous report.
E-mail messages include date and time stamps, permitting an objective
assessment of participants’ compliance with the instructions to complete a
Table 5
Study 3: Descriptive Statistics and Reliabilities
Group/variable MSDRange Cronbach’s
␣
Utility patentees
a
Inspiration (overall) 43.92 9.39 15–56 .93
Inspiration frequency 21.71 5.39 8–28 .94
Inspiration intensity 22.22 4.98 7–28 .92
Patents held (transformed) 0.38 0.40 0–.97
Sex 1.08 0.27 1–2
Age 50.53 12.41 27–87
Education level 4.57 1.90 1–7
Design patentees
b
Inspiration (overall) 41.05 11.53 8–56 .94
Inspiration frequency 20.11 6.44 4–28 .93
Inspiration intensity 20.95 5.73 4–28 .91
Patents held (transformed) 0.45 0.39 0–.99
Sex 1.24 0.43 1–2
Age 49.65 12.47 24–80
Education level 4.23 1.50 1–7
Alumni
c
Inspiration (overall) 34.07 11.49 8–56 .96
Inspiration frequency 16.49 6.21 4–28 .95
Inspiration intensity 17.58 5.88 4–28 .94
Sex 1.10 0.30 1–2
Age 49.97 15.30 22–82
Education level 5.95 0.84 5–7
a
n ⫽ 51.
b
n ⫽ 55.
c
n ⫽ 93.
880
THRASH AND ELLIOT

diary each night.
3
On the basis of the distribution of time stamps, 5:00 AM
was identified as a natural cut-off between days. E-mails received after
5:00 AM were considered late. Of the 2,100 possible daily diaries (i.e., 150
participants ⫻ 14 days), 2,079 (99%) were returned, and 1,816 (86%) were
returned on time. Given the high percentage of diaries returned on time,
late data were excluded from the analyses.
Measures
Inspiration
Trait inspiration was assessed with the IS.
Other Trait Measures
All 10 variables (excluding the OED measures) that correlated above .30
with the overall inspiration index in Study 2 (i.e., intrinsic motivation,
Openness to Experience, absorption, work mastery, creativity, positive
affect, perceived competence, self-esteem, optimism, self-determination)
were assessed again in this study by the instruments used previously. Given
that the PANAS item “inspired” loaded significantly on the inspiration
factors but not on positive affect in Study 2c, we used a nine-item measure
of positive affect that excluded this item.
Daily Measures
Each of the above variables was also assessed daily by single items.
Inspiration was assessed by the item “a feeling of inspiration,” intrinsic
motivation was assessed by “enjoyment,” Openness to Experience by
“openness to a new idea, behavior, feeling, etc.,” absorption by “absorption
in something I sensed or imagined,” work mastery by “a desire to master
something,” creativity by “creativity,” positive affect by “positive emo-
tion,” perceived competence by “competence/ability,” self-esteem by “a
positive view of myself,” optimism by “optimism,” and self-determination
by “freedom/choicefulness.” Participants rated each item using the follow-
ing scale: N ⫽ I did not experience this (coded as 0); 1 ⫽ I did experience
this, very mildly;2⫽ I did experience this, somewhat mildly;3⫽ I did
experience this, moderately;4⫽ I did experience this, somewhat intensely;
5 ⫽ I did experience this, very intensely. Analyses used both unaggregated
and aggregated data. Three types of aggregates were computed for each
variable. Overall daily aggregates were computed as the individual’s mean
for the variable across days. Following Schimmack and Diener (1997),
frequency aggregates were computed as the proportion of the individual’s
responses indicating that the experience was present (i.e., 1 or greater)
rather than absent (i.e., 0), and intensity aggregates were computed as the
mean response on those days in which the experience was present.
Results and Discussion
Predictive Validity
Zero-Order Predictive Validity
Descriptive statistics may be found in Table 7. The overall IS
index positively predicted the overall daily inspiration aggregate
(r ⫽ .38, p ⬍ .001). IS frequency positively predicted the inspi-
ration frequency aggregate (r ⫽ .31, p ⬍ .001) as well as the
inspiration intensity aggregate (r ⫽ .20, p ⬍ .05). IS intensity
positively predicted the inspiration intensity aggregate (r ⫽ .31,
p ⬍ .001) as well as the inspiration frequency aggregate (r ⫽ .33,
p ⬍ .001).
Discriminant Validity of Frequency and Intensity
Dimensions
We used a two-step procedure to examine the relations among
the unique components of the inspiration frequency and intensity
variables. First, we created two residualized variables, consisting
of the unique portions of the inspiration frequency and intensity
3
To protect against the possibility that participants might attempt to
falsify date and time stamps, we did not use the standard stamps that appear
in an E-mail header. Instead, we used the hidden (by default) time stamps
that were added to the message by the first server to receive it.
Table 6
Study 3: Fit Indices for Tests of Differences Across Groups
Model
␹
2
df CFI TLI RMSEA ⌬
␹
2
⌬df ⌬CFI
Tests of measurement invariance across groups
1. Baseline model: U ⫽ D ⫽ A 98.19*** 45 .97 .94 .08
2. Invariant loadings: U ⫽ D ⫽ A 108.33*** 57 .97 .96 .07
Model 2 vs. 1 10.13 12 .00
3. Invariant intercepts: U ⫽ D ⫽ A 119.01*** 69 .97 .96 .06
Model 3 vs. 2 10.68 12 .00
Tests of differences in construct parameters
4. Invariant latent variances: U ⫽ D ⫽ A 123.21*** 73 .97 .97 .06
Model 4 vs. 3 4.20 4 .00
5. Invariant latent covariance: U ⫽ D ⫽ A 126.19*** 75 .97 .97 .06
Model 5 vs. 4 2.98 2 .00
6. Invariant latent means: U ⫽ D 127.88*** 77 .97 .97 .06
Model 6 vs. 5 1.69 2 .00
7. Invariant latent means: U ⫽ D ⫽ A 154.90*** 79 .95 .95 .07
Model 7 vs. 6 27.02*** 2 ⫺.02
Note. The groups to which equality constraints were applied are indicated after each model description, where
U ⫽ utility patentees, D ⫽ design patentees, and A ⫽ alumni. Constraints were added cumulatively to previously
accepted models. The accepted model in each model comparison is indicated with italics. For all chi-squares,
N ⫽ 199. CFI ⫽ comparative fit index; TLI ⫽ Tucker–Lewis index; RMSEA ⫽ root-mean-square error of
approximation.
*** p ⬍ .001.
881
INSPIRATION

daily aggregates; second, we regressed these residuals on both IS
frequency and IS intensity. In these analyses, the inspiration fre-
quency residual was positively predicted (marginally) by IS fre-
quency, F(1, 145) ⫽ 2.84, p ⬍ .10 (
␤
⫽ .17), but not by IS
intensity (
␤
⫽ .06). The intensity residual was positively predicted
by IS intensity, F(1, 145) ⫽ 6.36, p ⬍ .05 (
␤
⫽ .26), but not by
IS frequency (
␤
⫽⫺.06).
Predictive Validity Controlling Measures of Other
Constructs
Three sets of analyses examined whether the IS indices predict
their corresponding daily aggregates while controlling correlates
of (a) the predictor, (b) the criterion, or (c) both the predictor and
the criterion. In the first set of analyses, overall daily inspiration,
daily frequency, or daily intensity was regressed on the corre-
sponding IS index and a trait correlate; 30 such analyses were
conducted (i.e., 3 inspiration indices ⫻ 10 correlates). In the
second set of analyses, the criterion was a daily inspiration variable
(overall inspiration, frequency, or intensity) from which the cor-
responding dimension of a daily correlate had been partialed; each
criterion variable was regressed on the corresponding IS subscale
(e.g., daily frequency of positive affect was partialed out of daily
frequency of inspiration, and the residual was regressed on IS
frequency). The third set of analyses controlled trait and daily
correlates simultaneously (e.g., daily frequency of positive affect
was partialed out of daily frequency of inspiration, and the residual
was regressed on IS frequency and trait positive affect). As shown
in Table 8, the IS indices predicted their corresponding daily
dimensions in each analysis.
Nomological Network (Revisited)
Inspiration’s nomological network (see Study 2) was reexam-
ined in several ways. Because the 10 correlate variables included
in this study had all correlated with both IS frequency and IS
intensity in Study 2, we focus here on overall inspiration for the
sake of parsimony and presentation clarity. As shown in the left
column of Table 9, the IS was positively related to all 10 of the
other trait measures, which thus replicates the results of Study 2.
To rule out the possibility that the obtained relationships were
artifacts of the trait-report methodology, we also computed corre-
lations using the aggregated daily measures of inspiration and
other constructs. As shown in the middle column of Table 9, these
analyses fully replicated the nomological network.
Both of the preceding analyses address the question of whether
individuals who experience high levels of inspiration are also high
on other constructs. Next we aim to extend the nomological
network to the state (within-person) level. In other words, across
days, do individuals’ departures from their own baseline in inspi-
ration covary with departures from baseline on other variables?
We computed within-person correlations separately for each indi-
vidual and combined them meta-analytically (see Reis, Sheldon,
Gable, Roscoe, & Ryan, 2000). As shown in the right column of
Table 9, inspiration was found to correlate with all 10 variables
across days.
Antecedents and Consequences
Our next step was to determine whether the variables that make
up inspiration’s nomological network also function as antecedents
or consequences and whether they do so in a meaningful way (e.g.,
we expected evocation-relevant variables to function as anteced-
ents and motivation variables as consequences). Using multilevel
modeling, we examined all 10 correlate variables as possible
antecedents and consequences at multiple levels of analysis. In all
models, day-level variables were centered around individuals’
Table 7
Study 4: Descriptive Statistics and Reliabilities
Variable MSDRange Cronbach’s
␣
Trait measures
a
Inspiration (overall) 38.51 7.87 8–56 .93
Inspiration frequency 19.17 4.43 4–28 .94
Inspiration intensity 19.35 4.28 4–28 .91
Intrinsic motivation 45.39 4.77 31–57 .71
Openness to Experience 41.73 5.96 25–55 .67
Absorption 53.79 7.40 34–68 .89
Work mastery 51.37 6.18 33–64 .75
Creativity 28.92 4.61 16–40 .79
Positive affect 33.27 4.42 18–45 .76
Perceived competence 32.60 5.53 15–45 .84
Self-esteem 32.01 5.69 12–40 .89
Optimism 21.21 4.05 10–29 .82
Self-determination 64.97 12.17 32–90 .86
Aggregated daily indices
(overall)
b
Inspiration 2.26 0.94 0.00–4.00 .86
Intrinsic motivation 3.57 0.82 1.00–5.00 .88
Openness to Experience 2.53 0.99 0.00–5.00 .88
Absorption 2.33 1.09 0.00–4.86 .91
Work mastery 2.90 1.11 0.00–5.00 .91
Creativity 2.07 1.08 0.00–5.00 .89
Positive affect 3.52 0.78 0.00–5.00 .86
Perceived competence 3.04 0.77 1.00–4.64 .89
Self-esteem 3.07 0.89 0.00–4.62 .89
Optimism 3.16 0.90 0.00–5.00 .85
Self-determination 2.83 0.96 0.00–5.00 .92
Aggregated daily
frequencies
b
Inspiration 0.78 0.24 0.00–1.00 .83
Intrinsic motivation 0.97 0.09 0.43–1.00 .85
Openness to Experience 0.83 0.21 0.00–1.00 .85
Absorption 0.78 0.27 0.00–1.00 .89
Work mastery 0.84 0.23 0.00–1.00 .88
Creativity 0.76 0.27 0.00–1.00 .86
Positive affect 0.96 0.12 0.00–1.00 .76
Perceived competence 0.95 0.11 0.55–1.00 .80
Self-esteem 0.93 0.15 0.00–1.00 .73
Optimism 0.93 0.15 0.00–1.00 .77
Self-determination 0.90 0.18 0.00–1.00 .91
Aggregated daily
intensities
c
Inspiration 2.83 0.63 1.00–4.00 .74
Intrinsic motivation 3.68 0.70 1.00–5.00 .81
Openness to Experience 3.00 0.76 1.00–5.00 .92
Absorption 2.93 0.78 1.00–4.86 .95
Work mastery 3.41 0.79 1.00–5.00 .90
Creativity 2.64 0.77 1.00–5.00 .93
Positive affect 3.64 0.63 1.50–5.00 .84
Perceived competence 3.19 0.64 1.00–4.64 .84
Self-esteem 3.23 0.71 1.00–4.62 .77
Optimism 3.37 0.69 1.50–5.00 .84
Self-determination 3.13 0.74 1.00–5.00 .92
a
n ⫽ 171.
b
n ⫽ 150.
c
n ⫽ 148 to 150.
882
THRASH AND ELLIOT

means, and person-level variables were centered around sample
means (Raudenbush & Bryk, 2002). The state–trait model exam-
ining antecedents was as follows (each candidate antecedent was
examined in a separate analysis):
I
S共j兲
⫽
␤
0
⫹
␤
1
共I
S共j⫺1兲
兲 ⫹
␤
2
共A
S共j⫺1兲
兲 ⫹ r (1a)
␤
0
⫽
␥
00
⫹
␥
01
共I
T
兲 ⫹
␥
02
共A
T
兲 ⫹ u
0
(1b)
␤
1
⫽
␥
10
⫹ u
1
(1c)
␤
2
⫽
␥
20
⫹ u
2
. (1d)
The dependent variable in Equation 1a, I
S(j)
, represents state
inspiration for a particular individual on a particular day j of the
diary, where j ranges from 2 to 14. Equation 1a specifies that state
inspiration varies as a function of two within-person variables—
the previous day’s state inspiration, I
S(j⫺1)
, and the previous day’s
state antecedent, A
S(j⫺1)
—as well as intercepts and slopes (
␤
0
,
␤
1
,
and
␤
2
) for each individual and a random error term (r). The
intercepts
␤
0
represent individuals’ mean levels of inspiration
across the 2 weeks of the diary. Equation 1b models each individ-
ual’s mean level of inspiration as a function of two between-
persons variables—trait inspiration as assessed by the IS (I
T
) and
the trait antecedent variable (A
T
)—as well as intercepts and slopes
(
␤
00
,
␤
01
, and
␤
02
) for the sample as a whole and a random error
term (u
0
). Equations 1c and 1d model the within-person slopes as
unconditional. In this model, a significant
␥
02
indicates that a trait
antecedent predicts individuals’ mean levels of inspiration across
the 2-week diary when trait inspiration is controlled. A significant
␤
2
indicates that individuals’ departures from their own inspiration
mean are predicted by the state antecedent variable on the previous
day when the previous day’s state inspiration is controlled.
If traits and states are viewed as extremes of a continuum, it is
possible that the antecedents emerge most clearly at some inter-
mediate level of analysis. Accordingly, we also examined the
following between-weeks antecedents model:
I
S共j兲
⫽
␤
0
⫹ r (2a)
␤
0
⫽
␥
00
⫹
␥
01
共I
W1
兲 ⫹
␥
02
共A
W1
兲 ⫹ u
0
. (2b)
where I
S(j)
represents inspiration for a particular individual on a
particular day j of the diary such that j ranges from 8 to 14 (i.e.,
Week 2), I
W1
represents an aggregate of Week 1 inspiration scores
(i.e., mean of Days 1–7), and A
W1
represents an aggregate of
Week 1 antecedent scores. In this model, a significant
␥
02
indicates
that the Week 1 antecedent predicts individuals’ mean levels of
inspiration during the second week when Week 1 levels of inspi-
ration are controlled.
The state–trait model examining consequences was as follows:
C
S共j兲
⫽
␤
0
⫹
␤
1
共I
S共j⫺1兲
兲 ⫹
␤
2
共C
S共j⫺1兲
兲 ⫹ r (3a)
␤
0
⫽
␥
00
⫹
␥
01
共I
T
兲 ⫹
␥
02
共C
T
兲 ⫹ u
0
(3b)
␤
1
⫽
␥
10
⫹ u
1
(3c)
␤
2
⫽
␥
20
⫹ u
2
. (3d)
This model is identical in form to the state–trait antecedents
model, but the dependent variable in Equation 3a is a candidate
Table 9
Study 4: Nomological Network of Inspiration (Between- and
Within-Person Correlations)
Variable
Between-persons Within-person
Trait
report
a
Aggregated
daily report
b
Unaggregated
daily report
b
Intrinsic motivation .36*** .35*** .28***
Openness to Experience .30*** .54*** .26***
Absorption .34*** .49*** .31***
Work mastery .36*** .50*** .22***
Creativity .33*** .58*** .34***
Positive affect .42*** .44*** .31***
Perceived competence .17* .60*** .32***
Self-esteem .17* .49*** .31***
Optimism .26** .56*** .39***
Self-determination .16* .49*** .38***
a
n ⫽ 171.
b
n ⫽ 150.
* p ⬍ .05. ** p ⬍ .01. *** p ⬍ .001.
Table 8
Study 4: Prediction of Daily Inspiration From Trait Inspiration
With Covariates Controlled
Control variable
Beta coefficient
Overall
inspiration Frequency Intensity
Trait variable controlled
Intrinsic motivation .39*** .29** .35***
Openness to Experience .40*** .34*** .31***
Absorption .35*** .29** .29***
Work mastery .33*** .28** .27**
Creativity .36*** .29*** .30***
Positive affect .33*** .26** .29**
Perceived competence .37*** .31*** .32***
Self-esteem .38*** .31*** .31***
Optimism .35*** .30*** .27**
Self-determination .38*** .32*** .32***
Daily aggregate controlled
Intrinsic motivation .36*** .29*** .29***
Openness to Experience .36*** .31*** .27**
Absorption .31*** .27** .26**
Work mastery .30*** .28*** .23**
Creativity .28*** .25** .24**
Positive affect .35*** .30*** .29***
Perceived competence .27** .23** .24**
Self-esteem .32*** .29*** .24**
Optimism .31*** .30*** .22**
Self-determination .30*** .27** .19*
Trait and daily variables controlled
simultaneously
Intrinsic motivation .37*** .27** .33***
Openness to Experience .37*** .32*** .27**
Absorption .32*** .28** .26**
Work mastery .27** .28** .18*
Creativity .30*** .25** .24**
Positive affect .35*** .27** .30***
Perceived competence .29*** .22** .26**
Self-esteem .33*** .29*** .26**
Optimism .33*** .30*** .24**
Self-determination .32*** .28** .20*
* p ⬍ .05. ** p ⬍ .01. *** p ⬍ .001.
883
INSPIRATION

consequence variable for a particular individual on a particular
day. A significant
␥
01
indicates that trait inspiration predicts indi
-
viduals’ mean levels of the consequence variable across the
2-week diary when trait levels of the consequence variable are
controlled. A significant
␤
1
indicates that individuals’ departures
from their own mean on the consequence variable are predicted by
the previous day’s inspiration when the previous day’s level of the
consequence variable is controlled.
Finally, we examined the following between-weeks conse-
quences model:
C
S共j兲
⫽
␤
0
⫹ r (4a)
␤
0
⫽
␥
00
⫹
␥
01
共I
W1
兲 ⫹
␥
02
共C
W1
兲 ⫹ u
0
. (4b)
In this model, a significant
␥
01
indicates that Week 1 inspiration
predicts individuals’ mean levels of the consequence variable in
Week 2 when the consequence variable at Week 1 is controlled.
We examined these models using HLM 5.04 (Raudenbush,
Bryk, & Congdon, 2000). Random error terms were retained when
significant (Kenny, Kashy, & Bolger, 1998). The relevant
between-persons and within-person (between-days) parameters
from the state–trait analyses are shown in the left and right col-
umns of Table 10, respectively, and the relevant parameters from
the between-weeks analyses are shown in the center column. In the
following, a variable is described as an antecedent or consequence
if it functioned in that way in at least one level of analysis. We then
discuss differences across the three levels of analysis.
The antecedent findings are consistent with our conceptualiza-
tion. Most notable is that Openness to Experience emerged as an
antecedent, whereas self-determination did not. These findings
suggest that inspiration is facilitated by receptiveness and that it
does not represent the active assertion of the self. Other anteced-
ents included positive affect and two of the resource variables,
optimism and self-esteem. It is likely that the positivity of these
variables facilitates breadth of attention and thinking (Fredrickson,
1998) and, thus, exposure to evocative influences. Finally, work
mastery and creativity also emerged as antecedents, which sug-
gests that inspiration tends not to involve pure passivity; rather, an
active engagement with the world of objects or ideas may facilitate
exposure to evocative influences (see also von Hartmann, 1884;
Wallas, 1926).
The consequence findings are also consistent with our concep-
tualization. Work mastery emerged as a consequence, which indi-
cates that inspiration has an enduring motivational impact. Ab-
sorption also emerged as a consequence, which suggests that
inspiration focuses attention on object qualities such as beauty
rather than promoting diffuse or unfocused arousal. Inspiration
was also found to lead to creativity and to enhanced levels of all
three resource variables (perceived competence, self-esteem, and
optimism), which suggests that inspiration facilitates transcen-
dence of constraints and enhances the self. Inspiration also led to
increased self-determination; thus, inspiration appears to involve
discovery rather than assertion of the self. Finally, Openness to
Experience also emerged as a consequence. Even though inspira-
tion involves a focusing of attention, it appears not to involve the
rigidity of focus that might be expected in the pursuit of lower
gratifications.
Results differed somewhat across the three levels of analysis.
Most notable is that the between-persons level revealed primarily
Table 10
Study 4: Time-Lagged Antecedents and Consequences of Inspiration at Three Levels of Analysis
Antecedents/consequences
Between persons
(traits predicting
2-week diary)
Between weeks
(Week 1 diary
predicting Week 2)
Between days
(Day j ⫺ 1 predicting
Day j)
Antecedents of inspiration
Intrinsic motivation ⫺.011 .150 .020
Openness to Experience ⫺.007 .240* .022
Absorption .016 .090 .036†
Work mastery .021* .188** ⫺.010
Creativity .016 .193* .042
Positive affect .024 .181† .055*
Perceived competence .011 .172 .003
Self-esteem .009 .221** .071*
Optimism .034* .191† .080**
Self-determination ⫺.003 .117 .047†
Consequences of inspiration
Intrinsic motivation .015† .072 ⫺.014
Openness to Experience .030** .170* .014
Absorption .018 .201** ⫺.003
Work mastery .031* .161* .011
Creativity .029** .141* .002
Positive affect .004 .075 ⫺.004
Perceived competence .024** .129* .009
Self-esteem .022** .041 ⫺.010
Optimism .015* .086 .016
Self-determination .028** .067 .020
Note. All parameters are unstandardized coefficients. N ⫽ 148.
† p ⬍ .10. * p ⬍ .05. ** p ⬍ .01.
884
THRASH AND ELLIOT

consequences, the between-weeks level revealed approximately
equal numbers of antecedents and consequences, and the between-
days level revealed only antecedents. Inspiration appears to be
more causally efficacious as a trait or chronic tendency than as a
state. However, state inspiration is likely to have consequences on
the day it occurs (rather than the next day), a hypothesis that must
await further research involving multiple assessments per day.
Nomological Network and Consequences Controlling
Positive Affect
One may wonder whether our results are attributable solely to
the variance that inspiration shares with positive affect. Thus, it
was important to demonstrate that inspiration’s nomological net-
work, or a portion of it, is related to inspiration independently of
positive affect. Similarly, the practical utility of the new construct
rests on its ability to predict outcomes beyond variance shared with
positive affect. We therefore reexamined inspiration’s nomological
network and consequences while controlling positive affect.
Nomological Network Controlling Positive Affect
Each correlation in Table 9 (excluding correlations with positive
affect) was recomputed as a partial correlation with positive affect
controlled. When we used trait report data, all but four correlates
remained significant with trait positive affect controlled: perceived
competence, self-esteem, optimism, and self-determination. When
we used aggregated daily data, only intrinsic motivation became
nonsignificant with aggregated positive affect controlled. When
we used within-person unaggregated daily data, all nine within-
person correlates remained significant with within-person positive
affect controlled. Overall, the correlates were quite robust when
positive affect was controlled. The correlates that were least inde-
pendent of positive affect (perceived competence, self-esteem,
optimism, self-determination, and intrinsic motivation) were those
that involve agency or enhancement of the self’s resources.
Consequences Controlling Positive Affect
Each consequence analysis reported in Table 10 was repeated
with initial positive affect as an additional simultaneous predictor.
Regarding between-persons consequences of IS trait inspiration,
five of the seven significant consequences remained significant
with trait positive affect controlled: Openness to Experience, work
mastery, creativity, perceived competence, and self-determination.
By comparison, positive affect predicted only two consequences in
this analysis: intrinsic motivation and perceived competence. Re-
garding between-weeks consequences, three of five remained sig-
nificant: Openness to Experience, absorption, and perceived com-
petence. Positive affect also predicted three consequences:
absorption, work mastery, and perceived competence. Regarding
between-days consequences, inspiration still did not predict any
variables. Positive affect significantly predicted one variable, self-
determination, but it was a negative predictor. Most of inspira-
tion’s positive consequences were thus robust when positive affect
was controlled, and, in fact, inspiration fared better than did
positive affect in accounting for these important outcomes.
In sum, Study 4 finds that (a) trait reports of inspiration predict
individuals’ daily experiences of inspiration, (b) the nomological
network holds both at the between-persons and at the within-
person levels, (c) inspiration has a meaningful pattern of anteced-
ents and consequences, and (d) the correlates and consequences of
inspiration are quite robust when positive affect is controlled.
GENERAL DISCUSSION
We have argued that the field of psychology has not given
adequate attention to inspiration. To facilitate research, we offered
a conceptualization of inspiration and conducted a series of studies
aimed at validating the construct and establishing its place in
empirical psychology. According to our conceptualization, inspi-
ration is characterized by evocation, motivation, and transcen-
dence. As a first step in our research, we developed the IS, a trait
measure. The IS was found to have strong psychometric proper-
ties: It was internally consistent (Studies 1a, 1b), it consisted of
distinguishable and internally consistent Frequency and Intensity
subscales (Studies 1a, 1b), it demonstrated measurement invari-
ance across time and populations (Studies 1b, 3), and it demon-
strated test–retest stability (Study 1b).
We established construct validity, in part, by documenting a
nomological network that is consistent with our conceptualization
(Studies 2a–2c). The IS was found to converge with a number of
motivation constructs, but only with those that are compatible with
evocation and transcendence. For instance, inspiration’s transcen-
dent quality implies approach rather than avoidance motivation. As
expected, the IS was found to correlate positively with BAS (but
not BIS), Extraversion (but not Neuroticism), and positive emo-
tionality (but not negative emotionality), variables that reflect an
underlying approach (vs. avoidance) temperament (Elliot &
Thrash, 2002). Positive affect, which has been widely posited to
play a role in approach motivation, was inspiration’s strongest
correlate; negative affect, in contrast, was unrelated to inspiration.
If one considers differentiated forms of approach motivation, one
would expect only some to be compatible with inspiration. Indeed,
inspiration was found to correlate positively with the work-
mastery component of need for achievement but negatively with
the competitiveness component, which reflects the typically mun-
dane (nontranscendent) desire to outperform others. Similarly,
inspiration was found to correlate positively with intrinsic moti-
vation but negatively with extrinsic motivation.
Nonmotivational components of the nomological network pro-
vide additional evidence for the evoked and transcendent nature of
inspiration. For instance, inspiration correlated positively with
Openness to Experience, which implies receptiveness to evocative
influences, but was unrelated to Conscientiousness, which implies
willfulness. Inspiration correlated positively with absorption, in
which attention is directed toward beauty or other object qualities,
and creativity, which represents transcendence of constraints in
one’s thinking and behavior. Inspiration was also found to corre-
late positively with perceived competence, self-esteem, and opti-
mism, consistent with our proposal that evocation, motivation, and
transcendence imply an influx or bolstering of psychological
resources.
Additional support for construct validity came from relation-
ships to variables outside the traditional trait domain. First, the IS
was found to converge with an unlabeled definition of inspiration
from the OED, which thus grounds the IS in a culturally shared
conception of inspiration that explicitly refers to evocation (e.g.,
885
INSPIRATION

infusion), motivation (e.g., impulse), and transcendence (e.g., ex-
alted). Second, inspiration was found to be unrelated to social
desirability, indicating that our results are not attributable to re-
sponse bias. Third, inspiration was positively related to the number
of majors that students were pursuing, suggesting that inspired
individuals are more engaged with their environment. Regarding
the content of students’ majors, inspiration was highest among
students majoring in the humanities. The humanities includes
disciplines such as art, religion, and philosophy, areas that are
often associated with inspiration and that are directly concerned
with transcendent values such as beauty, goodness, and truth.
Fourth, we documented a link between inspiration and an impor-
tant real-world criterion: U.S. patents. Patent holders were found to
experience considerably more inspiration than a comparison sam-
ple, and the frequency of their inspiration was related to the
number of patents held (Study 3). Finally, IS scores predicted
individuals’ ongoing, daily experiences of inspiration, even when
trait and daily covariates were controlled (Study 4).
Because we were interested in the inspiration construct per se
and not only in trait inspiration, we also examined the state of
inspiration (Study 4). Ten central components of inspiration’s
nomological network were found to extend from the trait
(between-persons) level to the state (within-person) level. In other
words, not only do individuals who differ in inspiration also differ
in their levels of the other constructs, but particular individuals’
day-to-day fluctuations in inspiration are related to their day-to-
day fluctuations in the other constructs. This finding links inspi-
ration in a dynamic way to the components of its nomological
network.
Next we conducted time-lagged antecedent and consequence
analyses to examine directional relationships between inspiration
and the components of its nomological network (Study 4). Anal-
yses were stringent in that initial levels of the dependent variables
were controlled in each analysis. Results were consistent with our
conceptualization. For instance, Openness to Experience func-
tioned as an antecedent, suggesting that Openness not only co-
varies with inspiration but facilitates it. Work mastery and absorp-
tion were documented as consequences, suggesting that inspiration
has an enduring impact on motivation and focus. Creativity and all
three resource variables (perceived competence, self-esteem, and
optimism) emerged as consequences, indicating that inspiration
facilitates transcendence of constraints and enhances the self. Two
other findings are particularly noteworthy. First, work mastery was
documented as an antecedent, supporting the conventional wisdom
(e.g., Wallas, 1926) that inspiration, although evoked, does not
involve pure passivity; rather, it favors the prepared mind. Second,
self-determination emerged as a consequence but not an anteced-
ent of inspiration, suggesting that the self cannot take credit for
inspiration; the self is discovered, not asserted.
Finally, we provided evidence that inspiration is not merely an
established construct repackaged. Study 2a found that inspiration
shares only 22% of its variance with the Big Five traits. In Study
2c, CFA indicated that trait inspiration is factorially distinct from
positive affect, its strongest correlate. In Study 4, most correlates
and consequences were found to be robust when variance associ-
ated with positive affect was controlled. In fact, trait inspiration
predicted 5 of 10 possible outcome variables (Openness to Expe-
rience, work mastery, creativity, perceived competence, and self-
determination) when both trait positive affect and trait measures of
the dependent measures were controlled, with a period of 3 weeks
between the trait assessments and the first outcome assessment.
Inspiration thus stands as an important empirical construct in its
own right, worthy of serious attention within mainstream psychol-
ogy. It is interesting that the components of inspiration’s nomo-
logical network that were least independent of positive affect were
those that involve agency or enhancement of resources. Positive
affect may contribute to the motivational aspect of inspiration,
providing sustenance and giving inspiration its dynamic character.
It is important to acknowledge a limitation of how we have
operationalized inspiration: In relying on the word inspiration and
its various forms, we have assumed that the lay public understands
the concept. The term inspiration has many shades of meaning,
and there was likely some variance in participants’ understanding
of the concept. However, this limitation may be more apparent
than real, for the following reasons. First, interview research by
Hart (1993, 1998) found that participants across populations have
fundamentally similar conceptions of inspiration: “Despite the
great many shades of meaning the term has in common usage it
appears to represent a clear and consistent event” (Hart, 1993, p.
144). Second, previous research relying on lay conceptions of
inspiration has yielded findings that are quite consistent with our
conceptualization (see the introduction). Third, we demonstrated
that the IS measure converges strongly with an unlabeled defini-
tion of inspiration from the OED. Fourth, we demonstrated a
stringent degree of measurement invariance across both time and
populations, tests that are sensitive to variance in how items are
interpreted (Little, 1997). Fifth, any variance in participants’ con-
ceptions of inspiration appears to have been inconsequential from
a practical perspective, in that our findings were consistent across
studies and highly supportive of our conceptualization. Finally, we
reiterate that our purpose is to learn about inspiration from the
participants who experience it. At this early stage of research, there
is much value in embracing lay conceptions as the gold standard
for empirical research, limitations notwithstanding.
A second limitation is that we have not examined what we might
call the internal structure of inspiration—for instance, how best to
categorize specific types of inspiration, whether they tend to co-
occur in the same individuals, and so on. As we demonstrated in
the introduction, the various types of inspiration share a common
set of core or genotypic features (evocation, motivation, and tran-
scendence), but there are, of course, differences across types. We
recommend that future research focus on both the general inspi-
ration construct and its specific manifestations, complementary
approaches that maximize bandwidth and fidelity, respectively.
Although the IS was developed as a general measure, researchers
could include directions that ask participants to answer with re-
spect to a particular content domain or type of inspiration.
In our research, we have linked inspiration to a number of
theoretical frameworks in personality and social psychology, such
as approach–avoidance motivation and the Big Five traits. In
closing, we place inspiration in the broader context of the field of
psychology. In the characteristics of evocation, motivation, and
transcendence, we see the contributions of Skinner, Freud, and
Maslow, respectively. Skinner taught psychologists to appreciate
the capacity of stimuli to evoke behavior. Freud revealed that
dynamic and compelling motivations may impinge on the ego.
Maslow showed how one may rise above ordinary desires and
participate in the transcendent. When one views the world through
886
THRASH AND ELLIOT

the lens of behaviorism, psychoanalysis, or humanism in isolation,
the topic of inspiration fails to present itself as relevant: Skinner
ruled out nonbehavioral constructs as unscientific, Freud failed to
acknowledge the positive aspects of human motivation, and
Maslow viewed transcendence as a quiescent state of being rather
than as motivational. However, when one views the world through
the lens of behaviorism, psychoanalysis, and humanism simulta-
neously, the inspiration construct comes into focus: The heights of
human motivation spring from the beauty and goodness that pre-
cede us and awaken us to better possibilities (see also Bradley,
1929; Dubay, 1999). Inspiration thus falls squarely within the
purview of psychology, and we hope that our research will serve
as an impetus for further study of this important and neglected
topic.
References
Allport, G. W., & Odbert, H. S. (1936). Trait-names: A psycho-lexical
study. Psychological Monographs, 47, 1–93.
Amabile, T. M., Hill, K. G., Hennessey, B. A., & Tighe, E. M. (1994). The
Work Preference Inventory: Assessing intrinsic and extrinsic motiva-
tional orientations. Journal of Personality and Social Psychology, 66,
950–967.
Andrews, F., & Withey, S. (1976). Social indicators of well-being. New
York: Plenum.
Aquinas, T. (1950). Summa theologica. Stanford, CA: Stanford University
Press.
Arbuckle, J. L. (1999). Amos 4. 01 [Computer software]. Chicago: Small-
Waters.
Averill, J. R. (1975). A semantic atlas of emotional concepts. Catalog of
Selected Documents in Psychology, 5, 330 (Manuscript No. 1103).
Bass, B. M., & Avolio, B. J. (1994). Improving organizational effective-
ness through transformational leadership. Thousand Oaks, CA: Sage.
Batson, C. D., Schoenrade, P., & Ventis, W. L. (1993). Religion and the
individual: A social-psychological perspective. New York: Oxford Uni-
versity Press.
Beer, R. D., Quinn, R. D., Chiel, H. J., & Ritzmann, R. E. (1997).
Biologically inspired approaches to robotics: What can we learn from
insects? Communications of the ACM, 40, 30–38.
Bollen, K. A. (1989). Structural equations with latent variables. New
York: Wiley.
Bowra, C. M. (1955). Inspiration and poetry. London: Macmillan.
Bradley, A. C. (1929). A miscellany. London: Macmillan.
Buber, M. (1996). I and thou. New York: Simon & Schuster.
Canale, F. L. (1994a). Revelation and inspiration: The classical model.
Andrews University Seminary Studies, 32, 7–28.
Canale, F. L. (1994b). Revelation and inspiration: The liberal model.
Andrews University Seminary Studies, 32, 169–195.
Carpenter, D. (1987). Inspiration. In M. Eliade (Ed.), The encyclopedia of
religion (pp. 256–259). New York: Macmillan.
Carver, C. S., & White, T. L. (1994). Behavioral inhibition, behavioral
activation, and affective responses to impending reward and punishment:
The BIS/BAS scales. Journal of Personality and Social Psychology, 67,
319–333.
Costa, P. T., & McCrae, R. R. (1992). Revised NEO Personality Inventory
(NEO–PI–R) and NEO Five Factor Inventory (NEO–FFI) professional
manual. Odessa, FL: Psychological Assessment Resources.
Crowne, D. P., & Marlowe, D. (1964). The approval motive. New York:
Wiley.
Dahl, H., & Stengel, B. (1978). A classification of emotion words: A
modification and partial test of de Rivera’s decision theory of emotions.
Psychoanalysis and Contemporary Thought, 1, 261–312.
Davitz, J. R. (1969). The language of emotion. New York: Academic Press.
Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and self-
determination in human behavior. New York: Plenum.
Dess, G. G., & Picken, J. C. (2000). Changing roles: Leadership in the 21st
century. Organizational Dynamics, 28, 18–33.
Dubay, T. (1999). The evidential power of beauty: Science and theology
meet. San Francisco: Ignatius Press.
Elliot, A. J. (1997). Integrating the “classic” and “contemporary” ap-
proaches to achievement motivation: A hierarchical model of approach
and avoidance achievement motivation. In M. L. Maehr & P. R. Pintrich
(Eds.), Advances in motivation and achievement (Vol. 10, pp. 143–179).
Greenwich, CT: JAI Press.
Elliot, A. J., & Church, M. A. (2001). [Fear of failure measures and their
correlates]. Unpublished raw data.
Elliot, A. J., & Thrash, T. M. (2002). Approach-avoidance motivation in
personality: Approach and avoidance temperaments and goals. Journal
of Personality and Social Psychology, 82, 804–818.
Epstein, S. (1994). Integration of the cognitive and the psychodynamic
unconscious. American Psychologist, 49, 709–724.
Fauteux, K. (1994). The recovery of self: Regression and redemption in
religious experience. New York: Paulist Press.
Finke, R. A. (1995). Creative insight and preinventive forms. In R. J.
Sternberg & J. E. Davidson (Eds.), The nature of insight. Cambridge,
MA: MIT Press.
Fredrickson, B. L. (1998). What good are positive emotions? Review of
General Psychology, 2, 300–319.
Frijda, N. H. (1994). Varieties of affect: Emotions and episodes, moods,
and sentiments. In P. Ekman & R. J. Davidson (Eds.), The nature of
emotion: Fundamental questions (pp. 59–67). New York: Oxford Uni-
versity Press.
Ghiselin, B. (Ed.). (1952). The creative process: A symposium. New York:
New American Library.
Gilson, E. (1953). Choir of muses. New York: Sheed and Ward.
Gnezda-Smith, N. (1994). The internal forces of creativity: When hearts
start to flutter. Roeper Review, 17, 138–143.
Gray, J. A. (1987). The psychology of fear and stress (2nd ed.). New York:
Cambridge University Press.
Haidt, J. (2000). The positive emotion of elevation. Prevention & Treat-
ment, 3.
Haidt, J. (2003). Elevation and the positive psychology of morality. In
C. L. M. Keyes & J. Haidt (Eds.), Flourishing: Positive psychology and
the life well-lived (pp. 275–289). Washington, DC: American Psycho-
logical Association.
Haidt, J., & Keltner, D. (in press). Awe/responsiveness to beauty and
excellence. In C. Peterson & M. E. P. Seligman (Eds.), The Values in
Action (VIA) classification of strengths. Cincinnati, OH: Values in
Action Institute.
Harding, R. E. M. (1948). An anatomy of inspiration. Cambridge, England:
Heffer.
Hart, T. R. (1993). Inspiration: An exploration of the experience and its
role in healthy functioning. Dissertation Abstracts International, 54(2-
B), 1077. (UMI No. 9316659)
Hart, T. (1998). Inspiration: Exploring the experience and its meaning.
Journal of Humanistic Psychology, 38, 7–35.
Hart, T. (2000). Inspiration as transpersonal knowing. In T. Hart, P. L.
Nelson, & K. Puhakka (Eds.), Transpersonal knowing: Exploring the
horizon of consciousness. Albany, NY: State University of New York
Press.
Harter, S. (1993). Visions of self: Beyond the me in the mirror. In J. E.
Jacobs (Ed.), Nebraska Symposium on Motivation: Vol. 40. Develop-
mental perspectives on motivation (pp. 99–144). Lincoln: University of
Nebraska Press.
Harvey, J. (1999). Music and inspiration. New York: Faber and Faber.
Helgeson, V. S., & Taylor, S. E. (1993). Social comparisons and adjust-
887
INSPIRATION

ment among cardiac patients. Journal of Applied Social Psychology, 23,
1171–1195.
Herman, W. E. (1990). Fear of failure as a distinctive personality trait
measure of test anxiety. Journal of Research and Development in Edu-
cation, 23, 180–185.
Hoyle, R. H. (1991). Evaluating measurement models in clinical research:
Covariance structure analysis of latent variable models of self-
conception. Journal of Consulting and Clinical Psychology, 59, 67–76.
Jung, J. (1986). How useful is the concept of role model?: A critical
analysis. Journal of Social Behavior and Personality, 1, 525–536.
Keltner, D., & Haidt, J. (in press). Approaching awe, a moral, spiritual, and
aesthetic emotion. Cognition and Emotion.
Kenny, D. A., Kashy, D. A., & Bolger, N. (1998). Data analysis in social
psychology. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), The
handbook of social psychology (4th ed., Vol. 1, pp. 233–265). New
York: McGraw-Hill.
Kris, E. (1952). Psychoanalytic explorations in art. New York: Schocken
Books.
Leavitt, J. (Ed.). (1997). Poetry and prophecy: The anthropology of inspi-
ration. Ann Arbor: University of Michigan Press.
Little, T. D. (1997). Mean and covariance structures (MACS) analyses of
cross-cultural data: Practical and theoretical issues. Multivariate Behav-
ioral Research, 32, 53–76.
Lockwood, P., & Kunda, Z. (1997). Superstars and me: Predicting the
impact of role models on the self. Journal of Personality and Social
Psychology, 73, 91–103.
Lockwood, P., & Kunda, Z. (1999). Increasing the salience of one’s best
selves can undermine inspiration by outstanding role models. Journal of
Personality and Social Psychology, 76, 214–228.
Lucas, R. E., Diener, E., Grob, A., Suh, E. M., & Shao, L. (2000).
Cross-cultural evidence for the fundamental features of extraversion.
Journal of Personality and Social Psychology, 79, 452–468.
Marsh, H. W. (1986). Global self-esteem: Its relations to specific facets of
self-concept and their importance. Journal of Personality and Social
Psychology, 51, 1224–1236.
Marsh, H. W., & O’Neill, R. (1984). Self description questionnaire III: The
construct validity of multidimensional self-concept ratings by late ado-
lescents. Journal of Educational Measurement, 21, 153–174.
Martindale, C. (1981). Cognition and consciousness. Homewood, IL:
Dorsey Press.
Maslow, A. H. (1968). Toward a psychology of being (2nd ed.). New York:
Van Nostrand Reinhold.
May, R. (1975). The courage to create. New York: Norton.
McAdams, D. P. (1982). Experiences of intimacy and power: Relationships
between social motives and autobiographical memory. Journal of Per-
sonality and Social Psychology, 42, 292–302.
McClelland, D. C., & Kirshnit, C. (1988). The effect of motivational
arousal through films on salivary immunoglobulin A. Psychology and
Health, 2, 31–52.
McCrae, R. R., & Costa, P. T. (1987). Validation of the five-factor model
of personality across instruments and observers. Journal of Personality
and Social Psychology, 52, 81–90.
McCrae, R. R., & Costa, P. T. (1997). Conceptions and correlates of
openness to experience. In R. Hogan, J. Johnson, & S. Briggs (Eds.),
Handbook of personality psychology (pp. 825–847). New York: Aca-
demic Press.
McCullough, M. E., Emmons, R. A., & Tsang, J.-A. (2002). The grateful
disposition: A conceptual and empirical topography. Journal of Person-
ality and Social Psychology, 82, 112–127.
McCutchan, A. (1999). The muse that sings: Composers speak about the
creative process. New York: Oxford University Press.
O’Brien, E. J., & Epstein, S. (1988). MSEI: The Multidimensional Self-
Esteem Inventory, professional manual. Odessa, FL: Psychological As-
sessment Resources.
Osborn, A. F. (1957). Applied imagination. New York: Scribner.
Pacini, R., & Epstein, S. (1999). The relation of rational and experiential
information processing styles to personality, basic beliefs, and the ratio-
bias phenomenon. Journal of Personality and Social Psychology, 76,
972–987.
Pleiss, M. K., & Feldhusen, J. F. (1995). Mentors, role models, and heroes
in the lives of gifted children. Educational Psychologist, 30, 159–169.
Preminger, A. (Ed). (1965). Inspiration. Encyclopedia of poetry and poetics
(pp. 396–398). Princeton, NJ: Princeton University Press.
Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models:
Applications and data analysis methods (2nd ed.). Thousand Oaks, CA:
Sage.
Raudenbush, S., Bryk, A., & Congdon, R. (2000). HLM for Windows
5.04 [Computer software]. Lincolnwood, IL: Scientific Software
International.
Raymond, G. L. (1907). The psychology of inspiration: An attempt to
distinguish religious from scientific truth and to harmonize Christianity
with modern thought. New York: Funk and Wagnalls.
Reis, H. T., Sheldon, K. M., Gable, S. L., Roscoe, J., & Ryan, R. M.
(2000). Daily well-being: The role of autonomy, competence, and relat-
edness. Personality and Social Psychology Bulletin, 26, 419–435.
Ribot, T. (1906). Essay on the creative imagination. Chicago: Open Court.
Robins, R. W., Hendin, H. M., & Trzesniewski, K. H. (2001). Measuring
global self-esteem: Construct validation of a single-item measure and the
Rosenberg self-esteem scale. Personality and Social Psychology Bulle-
tin, 27, 151–161.
Rogers, C. R. (1961). On becoming a person. Boston: Houghton Mifflin.
Rosenberg, M. (1965). Society and the adolescent self-image. Princeton,
NJ: Princeton University Press.
Rothenberg, A. (1990). Creativity and madness: New findings and old
stereotypes. Baltimore: Johns Hopkins University Press.
Rothenberg, A., & Hausman, C. R. (Eds.). (1976). The creativity question.
Durham, NC: Duke University Press.
Sadoski, M., & Paivio, A. (2001). Imagery and text: A dual coding theory
of reading and writing. Mahwah, NJ: Erlbaum.
Schachtel, E. G. (1959). Metamorphosis: On the development of affect,
perception, attention, and memory. New York: Basic Books.
Scheier, M. F., Carver, C. S., & Bridges, M. W. (1994). Distinguishing
optimism from neuroticism (and trait anxiety, self-mastery, and self-
esteem): A reevaluation of the Life Orientation Test. Journal of Person-
ality and Social Psychology, 67, 1063–1078.
Schimmack, U., & Diener, E. (1997). Affect intensity: Separating intensity
and frequency in repeatedly measured affect. Journal of Personality and
Social Psychology, 73, 1313–1329.
Schleiermacher, F. (1963). The Christian faith. New York: Harper & Row.
Schwo¨bel, C. (1987). Divine agency and providence. Modern Theology, 3,
225–244.
Sheldon, K. M., & Deci, E. L. (1996). The Self-Determination Scale.
Unpublished manuscript, University of Rochester.
Simpson, J. A., & Weiner, S. C. (Eds.). (1989). Oxford English Dictionary
(2nd ed., Vol. 7). Oxford, England: Clarendon Press.
Spence, J. T., & Helmreich, R. L. (1983). Achievement-related motives
and behaviors. In J. T. Spence (Ed.), Achievement and achievement
motives: Psychological and sociological approaches (pp. 7–74). San
Francisco: Freeman.
Steele, R. S. (1977). Power motivation, activation, and inspirational
speeches. Journal of Personality, 45, 53–64.
Sternberg, R. J., & Grigorenko, E. L. (2001). The (mis)organization of
psychology. APS Observer, 14, 1, 18.
Suler, J. R. (1980). Primary process thinking and creativity. Psychological
Bulletin, 88, 144–165.
Tabachnick, B. G., & Fidell, L. S. (2001). Using multivariate statistics (4th
ed.). Needham Heights, MA: Allyn & Bacon.
888
THRASH AND ELLIOT

Taylor, S. E., & Lobel, M. (1989). Social comparison activity under threat:
Downward evaluation and upward contacts. Psychological Review, 96,
569–575.
Tellegen, A. (1981). Practicing the two disciplines for relaxation and
enlightenment: Comment on “Role of the feedback signal in electro-
myograph biofeedback: The relevance of attention” by Qualls and Shee-
han. Journal of Experimental Psychology: General, 110, 217–226.
Tellegen, A. (1982). Brief manual for the Differential Personality
Questionnaire. Unpublished manuscript, University of Minnesota,
Minneapolis.
Thrash, T. M., & Elliot, A. J. (2001). Delimiting and integrating achieve-
ment motive and goal constructs. In A. Efklides, J. Kuhl, & R. M.
Sorrentino (Eds.), Trends and prospects in motivation research (pp.
3–21). Boston: Kluwer.
Tjas, K., Nelsen, E. A., & Taylor, M. (1997). Successful alumni as role
models for high school youth. High School Journal, 80, 103–110.
United States Patent and Trademark Office. (1999). Basic facts about
patents. Washington, DC: Author.
Vandenberg, R. J., & Lance, C. E. (2000). A review and synthesis of the
measurement invariance literature: Suggestions, practices, and recom-
mendations for organizational research. Organizational Research Meth-
ods, 3, 4–69.
von Hartmann, E. (1884). Philosophy of the unconscious. New York:
Macmillan.
Wallas, G. (1926). The art of thought. New York: Harcourt Brace.
Ward, T. B. (2001). Creative cognition, conceptual combination, and the
creative writing of Stephen R. Donaldson. American Psychologist, 56,
350–354.
Ward, T. B., Finke, R. A., & Smith, S. M. (1995). Creativity and the mind:
Discovering the genius within. New York: Plenum.
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and vali-
dation of brief measures of positive and negative affect: The PANAS
scales. Journal of Personality and Social Psychology, 54, 1063–1070.
Watson, D., Wiese, D., Vaidya, J., & Tellegen, A. (1999). The two general
activation systems of affect: Structural findings, evolutionary consider-
ations, and psychobiological evidence. Journal of Personality and Social
Psychology, 76, 820–838.
Widaman, K. F., & Reise, S. P. (1997). Exploring the measurement
invariance of psychological instruments: Applications in the substance
use domain. In K. J. Bryant, M. Windle, & S. G. West (Eds.), The
science of prevention: Methodological advances from alcohol and sub-
stance abuse research (pp. 281–324). Washington, DC: American Psy-
chological Association.
Wills, T. A., Windle, M., & Cleary, S. D. (1998). Temperament and
novelty seeking in adolescent substance use: Convergence of dimensions
of temperament with constructs from Cloninger’s theory. Journal of
Personality and Social Psychology, 74, 387–406.
Wood, J. V. (1989). Theory and research concerning social comparisons of
personal attributes. Psychological Bulletin, 106, 231–248.
Received March 19, 2002
Revision received August 27, 2002
Accepted September 30, 2002 䡲
Appendix
Inspiration Scale
Statement/item no. Statements and items Subscale
Statement 1 I experience inspiration.
Item 1f How often does this happen? Frequency
Item 1i How deeply or strongly (in general)? Intensity
Statement 2 Something I encounter or experience inspires me.
Item 2f How often does this happen? Frequency
Item 2i How deeply or strongly (in general)? Intensity
Statement 3 I am inspired to do something.
Item 3f How often does this happen? Frequency
Item 3i How deeply or strongly (in general)? Intensity
Statement 4 I feel inspired.
Item 4f How often does this happen? Frequency
Item 4i How deeply or strongly (in general)? Intensity
Note. The four Frequency items are rated on a scale from 1 (never)to7(very often). The four Intensity items
are rated on a scale from 1 (not at all)to7(very deeply or strongly). An f in item numbers indicates that the
item belongs to the Frequency subscale; an i indicates that it belongs to the Intensity subscale.
889
INSPIRATION